基于粒子群优化算法的电器工作状态辨识模型及其优化方法研究

Online identification on the health state parameters of thyristor modules in the HVdc converter is important to ensure the safe operation of the equipment. This article constructs a digital twin model of six-pulse converter by analyzing its working principle, the specific topology of thyristor modules in the valve assembly, and the available system information in practical engineering. The Runge–Kutta (RK) method is used to discretize the state equations of the six-pulse converter and to deduce the output data of digital twin model. The simulation results from a MATLAB/SIMULINK model of the six-pulse converter are utilized instead of the data from the actual physical model. According to the data from the simulated physical model and the digital twin model, the optimization objective function for the parameter identification is determined. Then, the equivalent insulation resistance and the damping capacitance parameters of each thyristor module are identified with the particle swarm optimization (PSO) algorithm. The results indicate that the identified equivalent insulation resistance and damping capacitance parameters have a maximum deviation of 7% from the true values. The identification errors of the damping capacitances are less than that of the equivalent insulation resistance, which is consistent with the trajectory sensitivity analysis. The contrast results show that the method in this article has better identification accuracy and noise immunity than those of the previous study. The proposed method provides a convenient solution for the intelligent maintenance of HVdc converter valves, without a large number of external sensors.

Localized discharge refers to the phenomenon of insulation material discharge in electrical equipment under high electric field strength due to uneven distribution of electric field. The occurrence of localized discharge in equipment poses significant risks to the insulation layer, and a rapid and accurate identification of discharge types is crucial for the normal operation of industrial systems. In regard to the problem of identifying localized discharge types in electrical equipment, considering the timeliness and accuracy requirements of the monitoring system, a method is proposed that involves constructing a phase distribution spectrogram of localized discharge to extract statistical features. The particle swarm optimization (PSO) algorithm is utilized to optimize a generalized regression neural network (GRNN) model. Finally, the statistical features are used as input to the neural network for discharge type identification. The results demonstrate that the proposed diagnostic method achieves high accuracy and efficiency in discharge type identification.

In order to realize the accurate recognition of vacuum degree of vacuum circuit breaker, a qualitative recognition method of terahertz spectrum based on PSO-SVM is proposed. Firstly, the time-domain spectral data of two vacuum degrees were obtained based on terahertz time-domain spectral system. Then, the two vacuum degrees were qualitatively identified by using support vector machine (SVM), genetic algorithm optimization support vector machine (GA-SVM) and particle swarm optimization support vector machine (PSO-SVM). The results show that the vacuum degree recognition model based on terahertz spectral data combined with PSO-SVM has the best effect, and the model accuracy rate is 94.1176%. Therefore, terahertz time-domain spectroscopy provides a new and efficient detection method for rapid qualitative identification of vacuum degree of vacuum circuit breakers, and also provides a practical basis for the application of terahertz technology in the field of non-destructive testing of electrical equipment.

The operation of electrical measuring equipment is affected by the local climate. Classifying the climate of each place based on environmental parameters is an important guide to the design and manufacture of electrical measurement equipment. Therefore, a support vector machine classification prediction model based on kernel parameter optimization is proposed. Firstly, linear discriminant analysis is used to extract the features of various environmental parameters. Then, support vector machine classification model is used to train the extracted environmental feature data. Finally, in order to improve the prediction accuracy of the model, particle swarm optimization algorithm was used to optimize the kernel parameters of the model to achieve the classification prediction of the model. The model is verified based on climate data, and the experimental results show that the use of the proposed model can effectively realize the classification and recognition of climate data.

DC microgrid with hydrogen-electric coupled characteristic is an effective solution for low-carbon development. To achieve the stable and efficient operation of different types of hydrogen and electrical equipment in the DC microgrid, this paper proposes a DC microgrid optimization control strategy based on parameter self-adaptive particle swarm optimization (PSA-PSO) algorithm. Firstly, this strategy takes economy as the optimization objective, and optimizes the operating state of hydrogen and electrical equipment with constraints of the equipment operation state and electricity balance. Then, in response to the problems existing in PSO algorithm, a PSA-PSO algorithm is proposed. This algorithm introduces Gaussian disturbance with adaptively adjusting the learning factor, inertia weight, and disturbance step size of the algorithm, thereby improving control performance. An example analysis was performed using real data, the results show that compared to traditional optimization control strategy, it can effectively improve the ability to search for optimal solutions, so that achieve adaptive coordinated optimization control of microgrid, and reduce operating cost.

Nowadays, the load monitoring system is an important element in smart buildings to reduce energy consumption. Nonintrusive load monitoring (NILM) is utilized to determine the power consumption of each appliance in smart homes. The main problem of NILM is how to separate each appliance's power from the signal of aggregated consumption. In this regard, this paper presents a combination between particle swarm optimization (PSO) and artificial neural networks (ANNs) to identify electrical appliances for demand‐side management. ANN is applied in NILM as a load identification task, and PSO is used to train the ANN algorithm. This combination enhances the NILM technique's accuracy, which is further verified by experiments on different datasets like Reference Energy Disaggregation Dataset, UK Domestic Appliance‐Level ElectricityUK‐DALE, and Indian data for Ambient Water and electricity Sensing. The high accuracy of the proposed algorithm is verified by comparisons with state of the art methods. Compared with other approaches, the total mean absolute error has decreased from 39.3566 to 18.607. Also, the normalized root mean square error (NRMSE) method has been used to compare the measured and predicted results. The NRMSE is in the range of 1.719%–16.514%, which means perfect consistency. This demonstrates the effectiveness of the proposed approach for home energy management. Furthermore, customer behavior has been studied, considering energy costs during day hours.

Cross-bonded cables improve transmission efficiency by optimizing the grounding method. However, due to the complexity of their grounding system, they are prone to multiple types of defects, making defect state identification more challenging. Additionally, accurately locating sheath damage defects becomes more difficult in cases of high transition resistance. To address these issues, this paper constructs a distributed parameter circuit model for cross-bonded cables and proposes a particle swarm optimization support vector machine (PSO-SVM) defect classification model based on the sheath voltage and current phase angle and amplitude characteristics. This model effectively classifies 25 types of grounding system states. Furthermore, for two types of defects—open joints and sheath damage short circuits—this paper proposes an accurate segment-based location method based on fault impedance characteristics, using zero-crossing problems to achieve efficient localization. The results show that the distributed parameter circuit model for cross-bonded cables is feasible for simulating electrical quantities, as confirmed by both simulation and real-world applications. The defect classification model achieves an accuracy of over 97%. Under low transition resistance, the defect localization accuracy exceeds 95.4%, and the localization performance is significantly improved under high transition resistance. Additionally, the defect localization method is more sensitive to variations in cable segment length and grounding resistance impedance but less affected by fluctuations in core voltage and current.

A large number of features are involved in fault diagnosis, and it is challenging to identify important and relative features for fault classification. Feature selection selects suitable features from the fault dataset to determine the root cause of the fault. Particle swarm optimization (PSO) has shown promising results in performing feature selection due to its promising search effectiveness and ease of implementation. However, most PSO-based feature selection approaches for fault diagnosis do not adequately take domain-specific a priori knowledge into account. In this study, we propose a correlation-guided PSO feature selection approach for fault diagnosis that focuses on improving the initialisation effectiveness, individual exploration ability, and population diversity. To be more specific, an initialisation strategy based on feature correlation is designed to enhance the quality of the initial population, while a probability individual updating mechanism is proposed to improve the exploitation ability. In addition, a sample shrinkage strategy is developed to enhance the ability to jump out of local optimal. Results on four public fault diagnosis datasets show that the proposed approach can select smaller feature subsets to achieve higher classification accuracy than other state-of-the-art feature selection methods in most cases. Furthermore, the effectiveness of the proposed approach is also verified by examining real-world fault diagnosis problems.

Intelligent fault diagnosis (IFD) plays a crucial role in reducing maintenance costs and enhancing the reliability of safety-critical energy systems (SCESs). In recent years, deep learning-based IFD methods have achieved high fault diagnosis accuracy extracting implicit higher-order correlations between features. However, the excessive long training time of deep learning models conflicts with the requirements of real-time analysis for IFD, hindering their further application in practical industrial environments. To address the aforementioned challenge, this paper proposes an innovative IFD method for SCES that combines the particle swarm optimization (PSO) algorithm and the ensemble broad learning system (EBLS). Specifically, the broad learning system (BLS), known for its low time complexity and high classification accuracy, is adopted as an alternative to deep learning for fault diagnosis in SCES. Furthermore, EBLS is designed to enhance model stability and classification accuracy with high-dimensional small samples by incorporating the random forest (RF) algorithm and an ensemble strategy into the traditional BLS framework. In order to reduce the computational cost of the EBLS, which is constrained by the selection of its hyperparameters, the PSO algorithm is employed to optimize the hyperparameters of the EBLS. Finally, the model is validated through simulated data from a complex nuclear power plant (NPP). Numerical experiments reveal that the proposed method significantly improved the diagnostic efficiency while maintaining high accuracy. In summary, the proposed approach shows great promise for boosting the capabilities of the IFD models for SCES.

No abstract available

This study addresses the issue of diagnosing faults in electric vehicle motors and presents a method utilizing Improved Wavelet Packet Decomposition (IWPD) combined with particle swarm optimization (PSO). Initially, the analysis focuses on common demagnetization faults, inter turn short circuit faults, and eccentricity faults of permanent magnet synchronous motors. The proposed approach involves the application of IWPD for extracting signal feature vectors, incorporating the energy spectrum scale, and extracting the feature vectors of the signal using the energy spectrum scale. Subsequently, a binary particle swarm optimization algorithm is employed to formulate strategies for updating particle velocity and position. Further optimization of the binary particle swarm algorithm using chaos theory and the simulated annealing algorithm results in the development of a motor fault diagnosis model based on the enhanced particle swarm optimization algorithm. The results demonstrate that the chaotic simulated annealing algorithm achieves the highest accuracy and recall rates, at 0.96 and 0.92, respectively. The model exhibits the highest fault accuracy rates on both the test and training sets, exceeding 98.2%, with a minimal loss function of 0.0035. Following extraction of fault signal feature vectors, the optimal fitness reaches 97.4%. In summary, the model constructed in this study demonstrates effective application in detecting faults in electric vehicle motors, holding significant implications for the advancement of the electric vehicle industry.

This paper proposes a novel fault diagnosis methodology for oil-immersed transformers to improve the diagnostic accuracy influenced by gas components in power transformer oil. Firstly, the Random Forest (RF) algorithm is utilized to evaluate and filter the raw data features, solving the problem of determining significant features in the dataset. Secondly, a multi-strategy Improved Particle Swarm Optimization (IPSO) is applied to optimize a double-hidden layer backpropagation neural network (BPNN), which overcomes the challenge of determining hyperparameters in the model. Four enhancement strategies, including SPM chaos mapping based on opposition-based learning, adaptive weight, spiral flight search, and crisscross strategies, are introduced based on traditional Particle Swarm Optimization (PSO) to enhance the model’s optimization capabilities. Lastly, AdaBoost is integrated to fortify the resilience of the IPSO-BP network. Ablation experiments demonstrate an enhanced convergence rate and model accuracy of IPSO. Case analysis using Dissolved Gas Analysis (DGA) samples compares the proposed IPSO–BP–AdaBoost model with other swarm intelligence optimization algorithms integrated with BPNN. The experimental findings highlight the superior diagnostic accuracy and classification performance of the IPSO–BP–AdaBoost model.

No abstract available

No abstract available

Compared to traditional power grids, microgrids have a more flexible operating mode. There are various distributed power sources within the microgrid, and different types of distributed power sources have different control methods. Once a short-circuit fault occurs in the microgrid, these characteristics will increase the difficulty of microgrid fault diagnosis and reduce the accuracy of microgrid fault diagnosis. This paper proposes an error-correcting particle swarm optimization back propagation microgrid fault diagnosis method for the diagnosis of short-circuit faults in microgrids that identifies the accuracy of alarm signals, corrects unreasonable signals, and obtains the correct fault set of the microgrid through the temporal logic relationship between each protection. Using the particle swarm optimization back propagation (PSO-BP) neural network algorithm to train fault alarm signals, fast convergence can be achieved, and accurate diagnostic results can be obtained after the sixth generation training is completed. As this fault diagnosis algorithm is applied to line protection equipment, it can be used to diagnose all types of short-circuit faults. This algorithm is easy to implement and has a small data scale, which is conducive to efficient and concise fault diagnoses in microgrids.

Safety and reliability are vital for robotic fish, which can be improved through fault diagnosis. In this study, a method for diagnosing sensor faults is proposed, which involves using Gramian angular field fusion with particle swarm optimization and lightweight AlexNet. Initially, one-dimensional time series sensor signals are converted into two-dimensional images using the Gramian angular field method with sliding window augmentation. Next, weighted fusion methods are employed to combine Gramian angular summation field images and Gramian angular difference field images, allowing for the full utilization of image information. Subsequently, a lightweight AlexNet is developed to extract features and classify fused images for fault diagnosis with fewer parameters and a shorter running time. To improve diagnosis accuracy, the particle swarm optimization algorithm is used to optimize the weighted fusion coefficient. The results indicate that the proposed method achieves a fault diagnosis accuracy of 99.72% when the weighted fusion coefficient is 0.276. These findings demonstrate the effectiveness of the proposed method for diagnosing depth sensor faults in robotic fish.

In response to the issue that the particle swarm optimization algorithm can only determine the minimum complete test set but cannot achieve optimal expected test cost in fault diagnosis problems, a new heuristic particle swarm optimization algorithm (HPSOIG) is designed by incorporating the information entropy algorithm into the heuristic function and fitness function. The algorithm performance is improved by combining randomness and directionality. Firstly, the particle swarm optimization algorithm is conceptually redefined, and the heuristic formula and fitness function in the heuristic particle swarm optimization algorithm are designed. Next, the mapping between the particle position and velocity in the particle swarm optimization algorithm and the test selection in the fault diagnosis problem is detailed, and the algorithm solving steps are introduced. Then, the running process of the algorithm is illustrated using a D matrix instance in the literature, and the accuracy and universality of the algorithm are verified through the radar system test D matrix constructed earlier and random simulation experiments. The results show that the HPSOIG algorithm solves the problem that the HPSO algorithm can only search for the minimum complete test set. Among the four compared algorithms, the HPSOIG algorithm can obtain the optimal expected test cost.

This article uses logistic chaotic mapping to improve the particle swarm algorithm parameters and construct the chaotic particle swarm optimization (CPSO) algorithm. Then, the CPSO algorithm is used to optimize the width, weight, and center values of the Radial Basis Function Neural Networks (RBFNN) to improve the RBFNN model used to diagnose transformer fault types. Results show that the CPSO-RBFNN model has a small mean square error and high accuracy in diagnosing transformer faults.

No abstract available

No abstract available

No abstract available

This study proposes an effective bearing fault diagnosis model based on an optimized approach for feature selection. The measured signal of the electric motor is processed by envelope analysis and Hilbert-Huang transform techniques to extract the potential features. An enhancement of the binary particle swarm optimization algorithm through population initialization strategy based on feature weights, new updating mechanism, and the screening and replacing process create a new and effective feature selection method that improves classification accuracy and reduces data size. The optimal feature subset is provided separately for artificial neural networks, and support vector machine classifier for the final recognition task. In multiple case studies, the proposed feature selection method is evaluated against the benchmark data sets and shows performance comparable to that of other peer competitors. The effectiveness of the proposed bearing fault diagnosis model is verified on different testbeds and achieves high accuracy and robustness under noise conditions. In addition, experimental results are compared with existing fault diagnostic models, showing the high possibility of the proposed bearing fault diagnosis model.

Intelligent motor fault diagnosis in industrial applications requires identifying key characteristics to differentiate various fault types effectively. Solely relying on statistical features cannot guarantee high classification accuracy, while complex feature extraction techniques can pose challenges for industry practitioners. Conversely, advanced feature extraction may not ensure that the model effectively learns these features for classification. A feature fusion approach that combines statistical and deep learning features to address these challenges is proposed. Since statistical features form the foundation for general feature extraction, statistical and deep learning features are combined using Extreme Gradient Boosting (XGBoost) algorithm with Particle Swarm Optimization (PSO). The PSO algorithm automates parameter tuning for XGBoost. A deep neural network (DNN) adaptively extracts hidden features, improving bearing fault classification precision using t‐SNE representation. Results successfully prove the DNN's ability to classify diverse motor faults using deep learning features. Thus, integrating statistical features with XGBoost further enhances DNN's performance. To ensure robustness, the proposed method has been compared with different motor fault classification methods and validated across different motor fault datasets, showcasing improved classification accuracy and robust performance, even amidst varying noise levels. This approach represents a promising advancement in intelligent fault diagnosis within industrial contexts.

No abstract available

Power transformer fault diagnosis exerts a vital part in the safe operation of power system. The PSO-SVM based on transformer fault diagnosis still has some shortcomings, such as slow convergence speed and easy to fall into local optimization. This dissertation proposes a transformer diagnosis method based on Improve Particle Swarm Optimization to support Vector Machine (MPSO-SVM). Adding disturbance to Particle swarm optimization (PSO) to disturb the position of such precocious particles, so as to get rid of local optimum. The case analysis represents that the diagnostic accuracy of MPSO-SVM is higher than that of PSO-SVM and Generalized Regression Neural Network (GRNN), and MPSO-SVM can effectively promote the fault diagnosis performance of transformer.

Imbalanced fault diagnosis becomes increasingly im-portant as the number of fault samples is relatively small in practical situations. Sparse auto-encoder(SAE) has been well addressed in fault diagnosis while it is not suitable for imbalanced fault diagnosis. Cost sensitive learning can be utilized to extend the sparse auto-encoder to cost sensitive sparse auto-encoder(CS-SAE). However, the class weights assigned to different classes are usually unknown in practice. So we propose to use particle swarm optimization(PSO) to optimize the class weights for cost sensitive sparse auto-encoder(PSO-CSSAE). The experiments have shown that the proposed approach consistently outperforms the state of the art on Tennesse Eastman(TE) dataset.

The safety and public health during nuclear power plant operation can be enhanced by accurately recognizing and diagnosing potential problems when a malfunction occurs. However, there are still obvious technological gaps in fault diagnosis applications, mainly because adopting a single fault diagnosis method may reduce fault diagnosis accuracy. In addition, some of the proposed solutions rely heavily on fault examples, which cannot fully cover future possible fault modes in nuclear plant operation. This paper presents the results of a research in hybrid fault diagnosis techniques that utilizes support vector machine (SVM) and improved particle swarm optimization (PSO) to perform further diagnosis on the basis of qualitative reasoning by knowledge-based preliminary diagnosis and sample data provided by an on-line simulation model. Further, SVM has relatively good classification ability with small samples compared to other machine learning methodologies. However, there are some challenges in the selection of hyper-parameters in SVM that warrants the adoption of intelligent optimization algorithms. Hence, the major contribution of this paper is to propose a hybrid fault diagnosis method with a comprehensive and reasonable design. Also, improved PSO combined with a variety of search strategies are achieved and compared with other current optimization algorithms. Simulation tests are used to verify the accuracy and interpretability of research findings presented in this paper, which would be beneficial for intelligent execution of nuclear power plant operation.

In order to make use of fewer fault data samples to diagnose the main fault types of circuit breakers accurately in real time, an intelligent fault diagnosis method for circuit breakers based on convolutional neural network (CNN) and quantum particle swarm optimization (QPSO) is proposed. Firstly, the key features of the circuit breaker operational signal are extracted through the CNN model, and the extracted feature vectors are input into the support vector machine (SVM) for fault diagnosis. In order to improve the diagnostic performance, this paper uses QPSO algorithm to optimize the parameters of the classifier, it effectively solves the local optimal problem. The experimental results show that the method presented in this paper has achieved good results in fault diagnosis of circuit breakers, and the accuracy of diagnosis is up to 100%, which highlights the superiority of this method.

Modern fault diagnosis faces an imbalance issue at both the sample and feature levels. A common approach is to combine balancing strategies with feature selection algorithms to address this problem. However, this may result in feature subsets that cannot accurately represent the true data distribution. This paper proposes an imbalanced fault diagnosis method called RPLFS-MOBPSO, which combines region purity (RP) with local feature selection to achieve class balance implicitly by partitioning local regions. RP is used as a new objective in the optimization process to achieve accurate fault detection. The adaptive reference point method from Non-Dominated Sorting Genetic Algorithm III is employed to maintain a diverse and evenly distributed external archive. On both the simulated Tennessee Eastma (TE) benchmark datasets and the real bearing experimental bench datasets, the proposed method achieves the best results compared to two LFS methods and four imbalanced ensemble algorithms based on different balancing strategies. Our source codes are available at: https://github.com/EMRGSZU/papers-code/tree/main/RPLFS-MOBPSO.

Bearing is one of the key components of a rotating machine. Hence, monitoring health condition of the bearing is of paramount importace. This paper develops a novel particle swarm optimization (PSO)-least squares wavelet support vector machine (PSO-LSWSVM) classifier, which is designed based on a combination between a PSO, a least squares procedure, and a new wavelet kernel function-based support vector machine (SVM), for bearing fault diagnosis. In this work, bearing fault classification is transformed into a pattern recognition problem, which consists of three stages of data processing. Firstly, a rich information dataset is built by extracting the features from the signals, which are decomposed by the nonlocal means (NLM) and empirical mode decomposition (EMD). Secondly, a minimum-redundancy maximum-relevance (mRMR) method is employed to determine a subset of feature that can provide an optimal performance. Thirdly, a novel classifier, namely LSWSVM, is proposed with the aid of a PSO, to provide higher classification accuracy. The key innovative science of this work is to propropose a new classifier with the aid of an new wavelet kernel type to increase the classification precision of bearing fault diagnosis. The merit features of the proposed approach are demonstrated based on a benchmark bearing dataset and a comprehensive comparison procedure.

This article presents an effective bearing fault diagnosis model based on multiple extraction and selection techniques. In multiple feature extraction, the discrete wavelet transform, envelope analysis, and fast Fourier transform are considered. While the combined binary particle swarm optimization with extended memory is focusing on feature selection. The current signals are analyzed by discrete wavelet transform. From there, the statistical features in the time and frequency domain are extracted by two techniques: envelope analysis, fast Fourier transform. Subsequently, the binary particle swarm optimization is combined with extended memory and two proposed position update mechanisms to eliminate redundant or irrelevant features to achieve the optimal feature subset. Besides, three classifiers including naïve Bayes, decision tree, and linear discriminant analysis are applied and compared to select the best model to detect the bearing fault.

No abstract available

Fault diagnosis of power transformer is indispensable for power system reliability. To improve the function of fault diagnosis and overcome the “code absence” problem of the traditional ratio method, this paper presents a novel approach for oil chromatographic fault diagnosis based on particle swarm optimization algorithm. The PSO algorithm is used to obtain the optimal three-ratio value that can best represent various fault types of power transformers, and then the change trend of the characteristic gas of the power transformer is analyzed to predict the possible faults. Combining the stereogram method and the optimal three-ratio method, A comprehensive fault diagnosis method for based on oil chromatographic distraction is obtained. In the end, simulation of the actual oil chromatographic data of the transformer verifies the accuracy and effectiveness of the proposed method.

With the development of social economy, the industrial field has entered the era of automation control. Due to the complexity and uncertainty in the industrial environment, electrical system failures are becoming increasingly frequent. In this paper, a common fault and the application of remote PLC detection combined with chaotic particle swarm optimization algorithm formula in electrical system fault can show the problems and solutions. However, the realization of remote PLC detection shows that necessary condition monitoring and fault diagnosis are carried out to ensure the life safety of staff and the reliable, normal and university operation of equipment. Finally, this article mainly analyzes the three types of data detected by remote PLC for electrical system faults

No abstract available

The assessment of the state of health of power semiconductors and the use of thermal observers rely on precise knowledge of the thermal impedance of the device, which is hard to monitor online with state-of-the-art approaches. This work proposes thermal digital twins (DTs), which create a real-time-capable digital replica of the physical thermal behavior and enable monitoring the thermal impedance online. The particle swarm optimization (PSO) algorithm and the dual extended Kalman filter (DEKF) are used to extract the thermal model for online monitoring. This is demonstrated for both approaches via a real-time simulation (RTS) where the reference chip temperature is given by a digital thermal model. A comparison of the approaches is given and the DEKF-based approach is chosen for the implementation of a multichip model with thermal cross-coupling. The convergence of the DEKF-based DTs is experimentally validated in the laboratory.

In three-phase dc/ac inverter, the power device is one of the most prone links to failure. Once it fails, it will cause serious accidents. Therefore, it is of great significance to carry out more accurate condition monitoring. In this paper, a digital twin method for IGBT parameter identification of a three-phase dc/ac inverter is proposed. The state monitoring of IGBT is emphatically studied, and the digital twin model of the three-phase inverter circuit is established. A particle swarm optimization algorithm is used to iterate the characteristic data, so that the digital twin model is gradually approximated with the physical circuit parameters. It is worth noting that by calculating the output part and core part of the digital twin model in order, the number of characteristic quantities identified at the same time is reduced, and the collected output current and voltage waveforms are screened, and different frequency quantities are selected to build a suitable target function. Through this method, the saturation voltage of IGBT, the forward voltage of the reverse parallel diode and the inductance value are accurately identified. Finally, a three-phase dc/ac inverter is built. Simulation and experimental results verify the effectiveness of this method.

Abstract Based on that, the Internet of Things (IoT) is used in industrial applications for monitoring and controlling various sensor operations. In existing work, IoT-based monitoring and controlling operations for industries are proposed. But, in this, real-time monitoring of the data is performed to take necessary actions. This may fail at a fraction of a second when the device crossed its breakpoint or threshold value. Hence, in this, an optimized deep learning approach (Convolutional Neural Network) is proposed for monitoring and controlling the temperature in electrical motors. Here, the controlling is performed by predicting the temperature using a deep learning approach. This helps to improve the controlling operations in the IoT environment and protect the device from Malfunctioning. The proposed approach is tested on the Kaggle Sensor dataset for electrical motors. The optimal hyperparameters for the CNN are determined through the hybrid particle swarm and genetic algorithm by minimizing the cost function. The cost function is to reduce the RMSE rate. This method’s presentation is evaluated and the fidelity of root mean square merits. The whole process is implemented using MATLAB R2020a version under Windows 10 environment.

No abstract available

The output device of the converter transformer is the connection structure between the high-voltage output line of the converter transformer and the current carrying terminal board in the high-voltage bushing oil. It plays an important role in ensuring the insulation performance of the converter transformer and bushing. The structure of the output device of the converter transformer is complex, with a composite insulation structure composed of liquid and solid, various irregular interfaces, and the effects of the DC voltage, AC voltage, and polarity reversal voltage. Therefore, the E-field distribution of the±800kV ultra-high voltage converter transformer output device is very complex. Therefore, the fiber optic sensing technology can be used to monitor the transformer oil near the outlet device. To ensure the normal operation of the fiber optic sensor, the optimized design field strength value of the outlet device is set to 26.1kV/mm. Based on simulation calculation results, multi-objective optimization method and particle swarm optimization algorithm are applied to optimize the design of the outlet device. Finally, the reasonable and feasible insulation structure design scheme for ±800kV ultra-high voltage converter transformer outlet device is proposed, which is structurally reasonable and engineering feasible.

This work employed the particle swarm optimization (PSO) algorithm to assess the trade-off between breakdown voltage (BV) and on-state resistance (<inline-formula> <tex-math notation="LaTeX">$\text{R}_{\textit {DS},\textit {on}}{)}$ </tex-math></inline-formula> in 4H–SiC metal oxide semiconductor field effect transistors (MOSFET) for power devices. In this work, the numerical model obtained after analyzing the resistance composition is utilized as the objective function in PSO to determine characteristic parameters in double-diffused metal oxide semiconductor field effect transistors (DMOSFET). These equations are input for the PSO algorithm. The derived characteristic parameters include the drift region doping concentration and thickness, cell size, channel length, JFET region length, JFET region thickness, and doping concentration. To adhere to common application constraints, this work optimizes these characteristic parameters to minimize the <inline-formula> <tex-math notation="LaTeX">$\text{R}_{\textit {DS},\textit {on}}$ </tex-math></inline-formula> under typical BV ranging from 100 to 2000 V. The <inline-formula> <tex-math notation="LaTeX">$\text{R}_{\textit {DS},\textit {on}}$ </tex-math></inline-formula> for some typical applications was extracted and validated through TCAD simulations to ensure algorithm accuracy. The reported results confirm that PSO yields superior outcomes and may be considered when designing devices. This work offers helpful insights into the design of characteristic parameters for 4H–SiC power DMOSFET devices and evaluates the feasibility of using PSO to optimize the characteristic parameters of power devices.

In our proposed approach, we integrate ADA boosting with particle swarm optimization-extreme learning machine (PSO-ELM) to enhance the accuracy of wind power estimation, addressing the inherent unpredictability and variability in wind energy. Initially, we refine the thresholds and input weights of the extreme learning machine (ELM) and then construct the PSO-ELM prediction model. ADA Boost is utilized to generate multiple weak predictors, each comprising a distinct hidden layer node. The PSO technique is then employed to optimize the input weights and thresholds for each weak predictor. The final forecast is attained by amalgamating and weighting the outcomes from each weak predictor using a robust wind power forecast model. Experimental validation utilizing data from Turkish wind turbines underscores the efficacy of our approach. Comparative analysis against contemporary techniques such as ensemble learning models and optimal neural networks reveals that our ADA-PSO-ELM model demonstrates superior accuracy and generalizability in predicting wind power output under real-world conditions. The proposed approach offers a promising framework for addressing the challenges associated with wind power estimation, thereby facilitating more reliable and efficient utilization of wind energy resources.

The nonlinear dielectric under transient voltage excitation can be equated to a three-branch parallel equivalent circuit with nonlinear resistors and nonlinear capacitors connected in series and parallel, It is a difficult problem to solve the parameters of the nonlinear device of the equivalent circuit, In this paper, we propose to solve this problem using particle swarm optimization improved by genetic algorithm. (The improved algorithm is referred to as ‘‘PSOGA’’). Firstly, a nonlinear dielectric three-branch parallel equivalent circuit is constructed, Deriving the nonlinear equation of the branch response current versus the transient excitation voltage, After that, the Simulink module in MATLAB is used to construct a simulation circuit for the transient dielectric properties of nonlinear dielectrics, Applying a decaying oscillation voltage with gradually increasing frequency and decreasing amplitude and obtaining the total response current, Finally, a particle swarm algorithm modified by genetic algorithm is written using MATLAB to solve for the equivalent parameters. The simulation results show that the particle swarm optimization improved by genetic algorithm can solve the equivalent circuit parameters of nonlinear dielectric transient dielectric properties.

In this paper, a novel outlier detection method is proposed for industrial data analysis based on the fuzzy C-means (FCM) algorithm. An adaptive switching randomly perturbed particle swarm optimization algorithm (ASRPPSO) is put forward to optimize the initial cluster centroids of the FCM algorithm. The superiority of the proposed ASRPPSO is demonstrated over five existing PSO algorithms on a series of benchmark functions. To illustrate its application potential, the proposed ASRPPSO-based FCM algorithm is exploited in the outlier detection problem for analyzing the real-world industrial data collected from a wire arc additive manufacturing pilot line in Sweden. Experimental results demonstrate that the proposed ASRPPSO-based FCM algorithm outperforms the standard FCM algorithm in detecting outliers of real-world industrial data. Note to Practitioners—Electric arc (which is governed by the current and arc voltage) plays a significant role in monitoring the operating status of the wire arc additive manufacturing (WAAM) process. The nominal periodic current and voltage may occasionally change abruptly due to anomalies (such as arc instability, unstable metal transfer, geometrical deviations, and surface contaminations), which would affect the quality of the fabricated component. This paper focuses on detecting possible anomalies by analyzing the current and voltage during the WAAM process. A novel clustering-based outlier detection method is proposed for anomaly detection where abnormal and normal instances are categorized into two separate clusters. A new particle swarm optimization algorithm is put forward to optimize the initial cluster centroid so as to improve the detection accuracy. The proposed outlier detection method is applied to real-world data collected from a WAAM pilot line for detecting abnormal instances. Experimental results demonstrate the effectiveness of the proposed outlier detection method. The proposed outlier detection method can be applied to other industrial applications including electrical engineering, mechanical engineering and medical engineering. In the future, we aim to develop an online outlier detection system based on the proposed method for real-time for anomaly detection and defect prediction.

No abstract available

The essence of NILM (Non-intrusive load monitoring) is load decomposition whose results can be further used to improve the energy acquisition system, intelligent power consumption system, and hold out two-way interactive service and intelligent power consumption service. In this paper, a PSO-SVM approach is proposed for load disaggregation in non-intrusive load monitoring. A method in view of the power difference is used to do event detection for switching state of electrical equipment which is proved to be very effective. A multi-feature classification (MFC) on account of PSO-SVM is suggested that it can recognize the switching state of electrical equipment. The simulation results showed that the accuracy rate can attain 95.3% for switched-on state and 96.2% for switched-off state which are more precise than GA-BP and GA-SVM.

The rising structural complexity of modern power systems present significant challenges for the reliable monitoring of insulation conditions in high-voltage electrical equipment. Transient high-frequency pulse signals, such as partial discharges (PDs), are key indicators in circuit anomaly detection. However, their weak amplitude and pronounced non-stationarity severely limit the effectiveness of traditional feature extraction methods. To overcome these challenges, this paper proposes a hybrid PSO-KELM model that couples time–frequency analysis with intelligent optimization. By applying the Short-Time Fourier Transform (STFT), the model accurately captures transient characteristics and resonance patterns in high-frequency pulses. A particle swarm optimization (PSO) algorithm with dynamic weight adjustment is designed to improve the contribution of key high-frequency features. A kernel-based extreme learning machine (KELM) is constructed for classification by combining nonlinear kernel mapping and regularization to improve the robustness of the classifier in high-noise environments. The simulation performance of the proposed method in a widely used benchmark shows higher accuracy in recognizing PDs signals compared to the conventional method, while maintaining high computational efficiency.

Fabric pilling performance is one of the key indicators for evaluating textile quality, but there is limited research on the effectiveness of pilling detection traceability and non-intrusive monitoring of detection equipment operating status. In this paper, a multi-source data-driven method for classifying the working status of fabric pilling performance detection is proposed. This study constructs a real-time non-intrusive monitoring system for fabric pilling detection in a laboratory environment, collecting multiple types of data such as electrical parameters of pilling detection equipment, personnel behavior, and equipment noise. GoogLeNet convolutional neural network is used to recognize high-dimensional audio data and achieve feature dimension reduction. By constructing a multi-classification algorithm based on Decision Tree-Support Vector Machine (DT-SVM) for the pilling detection process, a minimum accuracy of 95.62% is achieved in practical operation. This system not only perceives the relevant influencing factors of detection activities without interfering with normal detection activities but also effectively distinguishes various detection working states, providing new ideas for the effectiveness traceability of pilling detection activities.

The accurate prediction of battery state of charge (SOC) is one of the critical technologies for the safe operation of a power battery. Aiming at the problem of mine power battery SOC prediction, based on the comparative experiments and analysis of particle swarm optimization (PSO) and Categorical Boosting (Catboost) characteristics, the PSO-Catboost model is proposed to predict the SOC of a power lithium iron phosphate battery. Firstly, the classification model based on Catboost is constructed, and then the particle swarm algorithm is used to optimize the Catboost hyperparameters to build the optimal model. The experiment and comparison show that the optimized model’s prediction accuracy and average precision are superior to other comparative models. Compared with the Catboost model, the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) values of the PSO-Catboost model decreased by 12.4% and 25.4% during charging and decreased by 5.5% and 12.2% during discharging. Finally, the Random Forest (RF) and Extreme Gradient Boosting (XGBoost) models, both ensemble learning models, are selected and compared with PSO-Catboost after being optimized via PSO. The experimental results show that the proposed model has a better performance. In this paper, experiments show that the optimization model can select parameters more intelligently, reduce the error caused by artificial experience to adjust parameters, and have a better theoretical value and practical significance.

No abstract available

No abstract available

Stator electrical faults, including interturn faults and high-resistance connection faults, are common faults in permanent magnet synchronous motors. Conventional model-based diagnostic methods typically rely on precise parameter modeling and are restricted to specific control strategies, which limits their ability to achieve online real-time fault diagnosis and accurate fault classification. To overcome these issues, this article first analyzes the characteristics of voltage disturbances in the αβ-frame under fault conditions while considering various control strategies (e.g., id = 0, maximum torque per ampere, and flux weakening). Then, a highly adaptive voltage disturbance decomposition scheme is proposed by leveraging the disturbance observation capability of an extended state observer. The proposed scheme can separate the observed voltage disturbances into resistive and inductive differential-mode components effectively, that reflects the different physical behaviors between healthy and fault conditions clearly. Thereby, it can reduce dependency on precise modeling and overcome classification challenges. Finally, experimental results have verified that the proposed method exhibits high robustness and low computational complexity in online fault diagnosis and classification across diverse operating conditions, underscoring its potential for practical applications.

No abstract available

The Magnetic Barkhausen Noise (MBN) is a non-destructive testing method, which, due to its high sensitivity to changes in the microstructure of the material, is increasingly being applied with success as a tool for evaluation of magnetic material state and properties. However, it is no less difficult to analyze the measurement signals and their correct interpretation due to the complex, non-deterministic and stochastic nature of the Barkhausen phenomenon. Depending on the material to be examined, a signal with different characteristics can be observed. Frequently, a signal with multi-phase Barkhausen activity characteristics is obtained, like in the case of grain-oriented electrical steels. Due to the increased computational capabilities of computers, more and more advanced signal analysis methods are being used and artificial intelligence is being involved as well. Recently, the time–frequency (TF) approach for MBN signal analysis was introduced and discussed in several papers, where short-time Fourier Transform (STFT) found frequent application with promising results. Due to the automation of the search for diagnostic patterns, the stage of selecting transformation parameters becomes extremely important in the process of preparing training data for evaluation algorithms. This paper investigates the influence of the STFT computational window size on the material state evaluation results obtained using convolutional neural network (CNN). The studies were performed for MBN signals obtained from grain-oriented electrical steel with anisotropic properties. The carried out work made it possible to draw connections on the importance of the choice of the window during the implementation of CNN network training.

Electrical treeing, a significant degradation process in XLPE insulated cables, presents a substantial risk as it gradually weakens the insulation, ultimately leading to failure. Thus, understanding and identifying various stages of electrical tree growth is essential for implementing effective condition monitoring strategies. This study focuses on classification of electrical tree images into Inception, Propagation, and Breakdown stages utilizing state of the art CNN and pre-trained CNNs, including InceptionNet, ResNet, and EfficientNet. The impact of different optimizers such as Adam, RMSprop, and SGD on the performance of these approaches was evaluated in this study. The results showcased the effectiveness of the EfficientNet model with the RMSprop optimizer, achieving an accuracy of 98.78%.

Estimating the state-of-charge (SOC) of LiFePO4 (LFP) batteries is challenging due to their long open-circuit-voltage (OCV) plateau regions. In this paper, a SOC estimation method based on force-electrical coupled signals in the framework of filtering algorithm is designed by combining the first-order equivalent mechanical model (EMM) and the first-order RC equivalent circuit model (ECM). Combining the respective advantages of LFP battery force curve characteristics and OCV curve characteristics to achieve globally optimal SOC estimation with multi-sensor fusion. First, an innovative EMM is proposed to simulate the force variation of the battery with SOC change, and a particle swarm optimization (PSO) algorithm is used to identify the model parameters. Second, the ECM and EMM adopt a tandem structure to estimate the SOC using extended Kalman filtering (EKF), where a noise covariance adaptive updating step is introduced to dynamically adjust the estimation weights of the two models in order to address the problem that the non-monotonicity of the force-SOC curves may lead to erroneous convergence of the SOC. Finally, it is demonstrated that the root-mean-square error of the SOC estimation of this method is within 2%, which improves the accuracy by a factor of 2-3 compared to the traditional voltage feedback method.

This manuscript presents the design of a microgrid featuring solar and wind as uncontrollable energy sources, alongside controllable sources like batteries and a diesel generator, aiming to address power supply variations resulting from load fluctuations. Controllers are imperative to mitigate these challenges, and the manuscript emphasizes the need for precise tuning of gain values for optimal electrical energy utilization. In lieu of the trial-and-error approach, particle swarm optimization (PSO) is employed for enhanced steady-state response in the Microgrid. The study also introduces the application of proportional-integral (PI), proportional-integral-derivative (PID), and PID with feed forward (PIDF) controllers to effectively address and resolve identified issues ensuring improved system performance and consistent power supply stability in the microgrid system.

With the widespread application of unmanned aerial vehicles (UAVs) in fields such as aerial surveying, intelligent transportation, and military reconnaissance, the health status of their propulsion systems plays a critical role in ensuring flight safety. To address the difficulty of identifying minor faults and transitional states in UAV motors during actual operations, this paper proposes an intelligent multicondition fault diagnosis method based on Principal Component Analysis (PCA) and Particle Swarm Optimization-Support Vector Machine (PSO-SVM). Four typical operating conditions (normal, slight fault, transitional state, and severe fault) are simulated to construct sample datasets. Sixteen time-domain statistical features are extracted, and PCA is applied for dimensionality reduction. The PSO algorithm is introduced to automatically optimize the kernel parameters of the SVM, significantly improving the model's classification accuracy and generalization capability. Experiments are conducted using the T-Motor MN2212 motor, with vibration signals collected at a sampling rate of 2000 Hz. The constructed dataset is used to compare multiple models. Results show that the proposed model achieves a recognition accuracy of 98.06% on the test set, outperforming models such as KNN, RF, MLP, and AE-SVM. These findings validate the effectiveness and engineering potential of the proposed method in UAV propulsion system fault diagnosis.

Analog circuit is an crucial component of electronic equipment, and the ability to diagnose its fault state quickly and accurately is essential for ensuring the safety and reliability of these electronic equipment. This paper addresses the problems of low diagnostic accuracy and the difficulties associated with model parameter selection in traditional fault diagnosis methods, particularly when dealing with nonlinear and non-stationary fault signals. A fault diagnosis method for analog circuits is proposed, which utilizes Ensemble Empirical Mode Decomposition (EEMD) for features extraction, the Maximum Information Coefficient algorithm (MIC) for features selection, and Particle Swarm Optimization (PSO) for optimizing Support Vector Machine (SVM) classification. Firstly, EEMD is used to adaptively decompose the fault signals in the circuit to extract multi-scale fault features. Secondly, the extracted features are quantitatively evaluated using the Pearson correlation coefficient and energy value analysis, leading to the construction of a fault feature vector is constructed. On this basis, the MIC feature selection algorithm is applied to further optimize the feature vector. Finally, an efficient fault classification model is developed by optimizing the hyperparameters of the SVM model using PSO. The simulation results show that the proposed method effectively overcome the problems of model complexity and low classification accuracy caused by improper selection of wavelet basis function. The accuracy of fault diagnosis and the efficiency of model training are significantly superior to those of traditional methods.

The ZPW-2000K jointless track circuit is composed of the main track and the small track in the tuning area. Aiming at the complexity and randomness of the tuning area failure, an intelligent diagnosis model of support vector machine(SVM) based on particle swarm optimization(PSO) is proposed. Firstly, according to the structural composition of the tuning area, 10 voltage and current monitoring quantities in ZPW-2000K track circuit monitoring system of passenger dedicated line are selected to form fault data feature set. Secondly, particle swarm optimization is used in MATLAB to optimize the SVM model parameters, the SVM diagnosis model with optimal parameters is obtained, then carry out fault pattern recognition. Through simulation calculation, the prediction results of PSO-SVM model are compared with traditional SVM model and genetic algorithm optimized SVM model, which proves that the algorithm is a new method to effectively evaluate the type of fault diagnosis, and realizes 7 types of fault identification in the tuning area. The classification accuracy of the model can reach 95% and has good fault diagnosis ability.

In this paper, a Support Vector Machine (SVM) and the k nearest neighbor (KNN) classifiers are employed to diagnose the power transformer oil insulation using dissolved gas analysis (DGA). Different vectors are used as inputs for classifiers to achieve a maximum accuracy rate. The five input vectors that are considered: DGA in ppm, DGA in percentage, Dörnenberg ratios, Rogers ratios and Duval triangle reports. Concerning the classes of faults, five types are adopted as output of classifiers (PD, D1, D2, T1&T2 and T3). In order to perform a better fault classification, The SVM parameters have been optimized with a Particle Swarm Optimization (PSO). Using Duval triangle reports, the PSO-SVM algorithm provides the highest accuracy rate than the KNN algorithm one.

In view of the defects of low model performance of traditional support vector machine (SVM), firstly, a nonlinear inertia weight based on cosine function is used to balance the exploration and foraging ability of seagull optimization algorithm (SOA), then proposes to use reverse learning strategy to reduce the defect of local optimization of seagull population, and can obtains the best parameters of SVM. Then, this paper uses two benchmark functions to compare improved-SOA (ISOA), particle swarm optimization (PSO), and SOA optimization performance. Compared with SOA and PSO, ISOA has better optimization performance. Finally, based on ISOA-SVM, this paper uses three-fold cross validation to diagnose the DGA data, and the diagnostic results are compared with PSO-SVM and SOA-SVM, the diagnosis performance of this model is better.

To improve the problems such as low relevance, redundant features and the low accuracy of traditional fault diagnosis methods, a transformer fault diagnosis method based on the random forest (RF) feature optimization and improved sparrow search algorithm (ISSA)- support vector machine (SVM) was proposed. First, the transformer fault features were optimized through the RF-based average accuracy reduction method. Then, the traditional sparrow search algorithm (SSA) was improved using the Tent chaos mapping and Levy flight strategy, so that the support vector machine (SVM) parameters could be optimized. The proposed method was compared with the SVM, particle swarm algorithm-support vector machine (PSO-SVM), and sparrow search algorithm-support vector machine (SSA-SVM). The results show that the fault diagnosis rate of ISSA-SVM is 91.67%, which is 11.11%, 8.34% and 2.78% higher than that of SVM, PSO-SVM and SSA-SVM respectively. This result verified the effectiveness of the proposed method in the fault diagnosis of transformer.

In order to accurately identify the insulation and mechanical faults of the switchgear, a general recognition method based on the fusion of power frequency short-time energy and MFCC features is proposed to diagnose the fault of the running audio file of the switchgear. Firstly, the short-time energy features based on power frequency are extracted. Then the signal is converted in frequency domain to extract the MFCC (Mel frequency cepstrum coefficient), and the short-time energy and MFCC feature fusion matrix is constructed. Finally, based on SVM (support vector machine) classifier, PSO (particle swarm optimization) is used to optimize model parameters, and partial discharge and abnormal mechanical vibration of switchgear are diagnosed. In order to verify the effectiveness of the system, several groups of audio files collected under different working conditions are analyzed respectively. The results show that the proposed feature fusion and PSO-SVM method are superior to other algorithms, the recognition accuracy is above 92%. The system has low cost and strong practicability, greatly improves the inspection efficiency of power equipment, and provides a new idea for the audio fault detection of engineering field switchgear.

To address the situation of small transformer fault sample data and low accuracy, In this paper, quantum particle swarm optimization (QPSO) and multi-classification correlation vector machine (M-RVM) are combined with the fusion of multi-feature input information for transformer fault diagnosis. First, to compensate for the relatively single fault feature input, the data of the DGA technique, IEC tri-ratio method, and Doernenburg ratio method are feature preprocessed. Second, the QPSO algorithm is combined with the fault data for training to obtain the kernel parameters in M-RVM. Finally, ten feature inputs are added to the model for training, and the trained multi-classification correlation vector machine is used to perform fault diagnosis on the test data. After example and comparative experimental analysis, the diagnostic accuracy of the method is as high as 94.85%, which is 8.24%, 4.57%, 4.57%, and 3.15% higher than that of SVM, PSO-SVM, M-RVM, and PSO-RVM, respectively.

At present, the most widely used method of winding fault detection is frequency response analysis (FRA). To surmount the defects of slow convergence and low classification accuracy caused by inappropriate parameter selection of support vector machines (SVM), this paper proposes a transformer winding fault classification method based on bald eagle search (BES) algorithm to optimize the kernel parameter g and the penalty coefficient C in the SVM model, denoted as BES-SVM. The analysis of transformer winding faults shows that the BES- SVM model can diagnose faults more accurately. Compared with the traditional SVM and PSO-SVM methods, this method has higher diagnostic accuracy.

To address the poor detection and inspection issues of partial discharge (PD) in power equipment, this paper employs a feature fusion model based on a deep residual network and an attention mechanism to perform joint time-frequency feature extraction and dynamic weight assignment on multimodal data. Next, an improved support vector machine (SVM) combined with a particle swarm optimization (PSO) algorithm is used to classify and identify the fused features. During the inspection optimization phase, an inspection priority matrix is constructed based on historical fault data and equipment importance levels. This method achieves an AUC value exceeding 0.974 for four typical discharge types in real-world power scenarios. Inspection optimization utilizes an ant colony algorithm to dynamically plan paths, increasing the detection rate of high-risk equipment to 100% and shortening path lengths. This paper provides reliable technical support for power equipment condition assessment.

No abstract available

Shipboard power systems are the core of maritime electrical automation, playing a crucial role in ensuring the normal operation of vessels. Efficient fault diagnosis facilitates rapid localization and judgment of faults, thereby safeguarding the safety of maritime operations. This research uses Simulink to simulate and analyze line faults in shipboard area distribution power systems. The simulation model is constructed to emulate common line faults in the power system. The study employs the Empirical Mode Decomposition (EMD) algorithm and Hilbert-Huang Transform as methods for extracting fault characteristics in shipboard power distribution systems. EMD is utilized to analyze the characteristic quantities of voltage and current fault signals, followed by the extraction of fault features from intrinsic mode functions through the Hilbert-Huang Transform. The extracted fault feature vectors serve as input parameters for a convolutional neural network (CNN) optimized using the particle swarm optimization (PSO) and support vector machine (SVM) to establish a fault diagnosis model. The analysis demonstrates that the PSO-SVM-CNN model significantly enhances the fault diagnosis accuracy of shipboard area power distribution systems, achieving a precision rate of up to 99.5%. This research contributes to advancing the safety and reliability of shipboard electrical systems through effective fault diagnosis methodologies.

When a series arc fault occurs in indoor power distribution system, current value of circuit is often less than the threshold of the circuit breaker, but the temperature of arc combustion can be as high as thousands of degrees, which can lead to electrical fire. The arc fault experimental platform is used to collect circuit current data of normal work and arc fault. Five types of loads which are commonly used in indoor distribution system, such as resistive and inductive in series load, resistive load, series motor load, switching power load and eddy current load, are chosen. This paper uses four features of current in time domain, i.e. current average, current pole difference, difference current average and current variance. Ten features of current in frequency domain are extracted by Fast Fourier Transform (FFT). The learning vector quantization neural network (LVQ-NN) is designed to judge the load type. The support vector machine optimized by particle swarm optimization (PSO-SVM) is designed to detect the arc fault. The simulation results show the effectiveness of the proposed method.

No abstract available

Fast and accurate fault classification is essential to power system operations. In this paper, in order to classify electrical faults in radial distribution systems, a particle swarm optimization (PSO) based support vector machine (SVM) classifier has been proposed. The proposed PSO based SVM classifier is able to select appropriate input features and optimize SVM parameters to increase classification accuracy. Further, a time-domain reflectometry (TDR) method with a pseudorandom binary sequence (PRBS) stimulus has been used to generate a dataset for purposes of classification. The proposed technique has been tested on a typical radial distribution network to identify ten different types of faults considering 12 given input features generated by using Simulink software and MATLAB Toolbox. The success rate of the SVM classifier is over 97%, which demonstrates the effectiveness and high efficiency of the developed method.

No abstract available

The operation of a circuit breaker involves control of secondary electrical circuits and transfer of energy among mechanical parts. The resulting vibration and current signals carry rich condition information. This paper proposes a diagnosis method for high-voltage circuit breakers based on multiscale sample entropy and current features. First, vibration signals are processed by wavelet-packet denoising and their multiscale sample entropy is computed. Then the current is filtered by a Butterworth filter and time-domain and statistical features are extracted to form an electro-vibration multivariate feature vector. Finally, a support vector machine (SVM) optimized by particle swarm optimization (PSO) is used to diagnose mechanical faults. An experimental platform is built with a ZN-65 vacuum circuit breaker. The proposed method achieves 94 % accuracy with low computational time and shows promise for engineering use.

No abstract available

Due to the complexity of the working environment of wind power generation systems, wind turbine power converters (WPC) can experience different types of faults. Traditional fault diagnosis methods suffer from issues such as the need for additional hardware, low accuracy, long execution time, and applicability only to small sample offline fault diagnosis. In order to address these problems, this article proposes a particle swarm optimization-based support vector machine (SVM) algorithm. The algorithm combines PSO algorithm, SVM algorithm, and moving average algorithm to effectively improve the robustness and accuracy of the fault diagnosis algorithm, while reducing the execution time and cost. This article selects three-phase current signals and bus voltage signals as fault diagnosis data, and then uses the moving average algorithm to process the fault data of the power converter, retaining the data features based on effectively smoothing the data. Finally, an improved particle swarm algorithm is used to construct a fault diagnosis model based on support vector machines for diagnosing open circuit faults in the power converter. In a dataset containing 9800 training samples and 4200 testing samples, the accuracy of the training samples is 98.898%, and the accuracy of the testing samples is 98.4524%. This effectively solves the problem of traditional SVM methods being only able to handle small batches of nonlinear datasets. Finally, this article compares the proposed fault diagnosis method with other types and similar types of fault diagnosis methods, verifying the effectiveness and superiority of the proposed approach.

Wireless power transfer (WPT) systems may encounter issues such as device short circuits, open circuits, and parameter drifts during operation. If these faults are not promptly diagnosed and resolved, they can severely impact the system. This paper focuses on the research of fault diagnosis techniques for WPT systems. By using parameters such as current and voltage within the system, the particle swarm optimization (PSO) algorithm optimizes the support vector machine (SVM) model parameters, constructing a high-accuracy fault diagnosis model. A LCC-LCC type WPT system was built for simulation and experimental verification. The results demonstrate that the proposed method can achieve accurate fault diagnosis, effectively ensuring the safety of WPT systems.

To address the challenge that faults in central air-conditioning chillers are difficult to detect in time and the types of faults are easily misidentified, an ensemble fault diagnosis model based on AdaBoost-PSO-SVM is proposed in this paper. Firstly, the Support Vector Machine (SVM) model, which is widely applied to nonlinear problems, is chosen as the foundational model for fault diagnosis. Subsequently, the Particle Swarm Optimization algorithm (PSO) is used to find the best of the penalty factor C and the kernel parameter g of the SVM within a given range, resulting in the creation of the base learner PSO-SVM of the ensemble model. Finally, the AdaBoost-PSO-SVM fault diagnosis model integrates multiple PSO-SVM learners by updating the sample weights and assigning different weights to the base learners. The effectiveness of the proposed AdaBoost-PSO-SVM model is evaluated using experimental data from ASHRAE RP-1043. Results demonstrate its ability to accurately identify chiller faults, even including low severity levels, with an average diagnosis accuracy rate of 90.69%.

Equipment failures often occur in the internal insulation of power transformers, mainly due to insulation aging, poor operating conditions, and unqualified manufacturing quality of power transformers. The insulating ability of the outer layer oil paper of the transformer winding is crucial, and the key factor affecting oil paper insulation is the hot spot fault of the transformer winding. Therefore, hot spot fault detection of transformers in operation is of great significance. At present, the existing hot spot temperature measurement methods at home and abroad are difficult to apply to transformers that have already been put into operation. Therefore, this paper proposed a transformer hot spot fault diagnosis method based on the combination of ultrasonic sensing technology and PSO-SVM algorithm. The three-dimensional model of the transformer was built using COMSOL finite element simulation software to simulate the dynamic response of ultrasonic signals received by sensors under different hot spot types, and then feature parameters were extracted based on the simulation data. Finally, the particle swarm optimization (PSO) and support vector machine (SVM) algorithm were combined to achieve the transformer hot spot temperature retrieval and fault location. The results indicated that the PSO-SVM model could effectively identify hot spot faults and achieve non-destructive online monitoring of internal hot spots in transformers, providing a new approach for the field of transformer hot spot fault diagnosis.

This study aims to improve the real-time monitoring and fault diagnosis of distribution transformers by utilizing a combination of five thin film gas detectors, these detectors include metal-modified graphene composite films and SnO2/RGO humidity sensors, which were prepared using the hydrothermal method. The experiment focused on investigating humidity and main fault characteristic gases that can reflect the insulation status of transformers. Additionally, a gas sensor array was constructed using a deep confidence neural network model. Based on the analysis of dissolved gas in transformer oil, the study extensively discusses the insulation fault diagnosis model and constructs the transformer fault diagnosis model using various methods including TRM, Particle swarm optimization support vector machine. The results demonstrated that the SnO2/RGO thin film humidity sensor exhibited high humidity sensitivity, and the other thin film gas sensors also exhibited good sensitivity. The average accuracy of the three classification methods mentioned is 80%, 92%, and 96%, respectively. These findings highlighted that the vector machine model not only improved the fault diagnosis accuracy but also possessed the characteristics of fewer parameters and a fast rate of convergence. Consequently, it effectively addressed the issue of early diagnosis of potential transformer faults. This study was of significant practical importance for ensuring the secure operation of the power grid.

Aimed at the power switch open circuit fault occurred in the inverter, the paper proposed one dimension convolutional autoencoder (1DCAE), particle swarm optimization (PSO) and support vector machine (SVM) based fault diagnosis algorithm of inverter. 1DCAE is used to obtain fault feature from phase current of inverter, PSO is used to determine penalty parameter C and kernel function parameter with SVM. The experimental results show that the techniques proposed in the paper has higher fault diagnosis accuracy.

The goal of the paper is to present a solution to improve the fault detection accuracy of rolling bearings. The method is based on variational mode decomposition (VMD), multiscale permutation entropy (MPE) and the particle swarm optimization-based support vector machine (PSO-SVM). Firstly, the original bearing vibration signal is decomposed into several intrinsic mode functions (IMF) by using the VMD method, and the feature energy ratio (FER) criterion is introduced to reconstruct the bearing vibration signal. Secondly, the multiscale permutation entropy of the reconstructed signal is calculated to construct multidimensional feature vectors. Finally, the constructed multidimensional feature vector is fed into the PSO-SVM classification model for automatic identification of different fault patterns of the rolling bearing. Two experimental cases are adopted to validate the effectiveness of the proposed method. Experimental results show that the proposed method can achieve a higher identification accuracy compared with some similar available methods (e.g., variational mode decomposition-based multiscale sample entropy (VMD-MSE), variational mode decomposition-based multiscale fuzzy entropy (VMD-MFE), empirical mode decomposition-based multiscale permutation entropy (EMD-MPE) and wavelet transform-based multiscale permutation entropy (WT-MPE)).

Traditional fault diagnosis methods of rotating machinery primarily rely on signal preprocessing, feature extraction, and classification. For the issue of small sample data, in this paper, an innovative adaptive and intelligent method for mechanical fault diagnosis is introduced to obtain more significant signal characteristics and improve recognition accuracy. First, variational mode decomposition(VMD) algorithm which established by Genetic Algorithm(GA) is used to decompose the original vibration signal into several intrinsic mode functions(IMFs). Seconedly, calculate the multi-scale permutation entropy(MPE) values of each IMF component to further extract features and construct feature vectors as inputs for the classifier. Finally, the optimal support vector machine (SVM) classifier was established through particle swarm optimization (PSO) to achieve fault diagnosis of rotating machinery. The experiment shows that the recognition rate of 97.7% is achieved by the proposed method.

This study aims to develop a real-time model for detecting and classifying various fault types in power distribution networks to enhance reliability and operational efficiency. A hybrid DWT–SVM approach is essential for accurately detecting and classifying faults in modern power distribution systems, especially under non-stationary and dynamic conditions. The proposed method addresses a range of fault scenarios, including single-phase-to-ground, line-to-line, double-line-to-ground, and three-phase faults, within a 33 kV distribution network. A hybrid approach combining Discrete Wavelet Transform (DWT) and Support Vector Machine (SVM) is introduced. Using the Debauchies-4 (Db4) wavelet, DWT effectively decomposes transient fault currents at the source terminal, capturing critical time-frequency domain features. Fault classification is performed using an SVM optimized with a Radial Basis Function (RBF) kernel and Particle Swarm Optimization (PSO), enabling precise mapping of data into higher-dimensional spaces for optimal separation. Validation conducted with MATLAB R2023b demonstrates a detection accuracy of 100% and a classification accuracy of 99%. Comparative analyses against models such as PSO-based Support Vector Machine (PSO-SVM), DWT-DNN and wavelet transform with artificial neural networks (WT-ANN) on the IEEE 13-bus system highlight the proposed method's superior performance. This innovative approach proves to be robust, adaptable, and highly effective across diverse fault scenarios, offering significant improvements in accuracy and reliability for fault detection and classification in power distribution networks. The method supports real-time operation by using high-speed simulation platforms MATLAB/Simulink to process signals and classify faults within milliseconds using a pre-trained SVM model.

The prominent vibration characteristics of amorphous alloy transformer (AMT) make it possible to apply the vibration method for real-time fault monitoring of AMT. Therefore, in order to solve the AMT vibration monitoring problem and enhance the diagnostic efficiency, this study proposes an AMT fault diagnosis model based on particle swarm optimization (PSO) to optimize the parameters of wavelet packet transform (WPT) and support vector machine (SVM).The optimal vibration signal acquisition point is determined by finite element analysis to ensure high signal quality. The PSO algorithm is used to optimize the number of WPT decomposition layers, wavelet basis functions, SVM kernel parameters g and penalty parameters c to enhance the accuracy of feature extraction and classification. Additionally, principal component analysis (PCA) reduces the dimensionality of the redundant frequency band energy after WPT feature extraction, minimizing data redundancy. Overall, the full-process optimization significantly improves AMT fault diagnosis efficiency compared with single-aspect optimization.

No abstract available

No abstract available

No abstract available

The anomaly identification research is carried out for the anomalous data existing in the load and power consumption dataset. Firstly, the LSSVM algorithm is used to identify the abnormal data, in view of the problem of low recognition accuracy of the LSSVM algorithm, the LSSVM algorithm optimized based on the improved PSO algorithm is proposed; then, the PSO-LSSVM algorithm is trained by using the historical load and power consumption dataset and abnormal dataset, and the abnormal data recognition model is obtained to identify the daily load curve which contains the abnormal data of load and power consumption. Finally, the feature curve judgment method is used to determine the exact location of the abnormal load and power consumption data in the curve. Comparison and analysis of simulation examples verify the reliability of PSO-LSSVM algorithm in identifying the abnormal electricity data.

Catenary has been exposed outdoors for a long term, and its failure rate is very high, which has seriously affected the operation and development of traction power supply system. Due to the problems of long detection time, backward detection means and influenced by human factors in traditional catenary fault identification methods, this paper proposed a fault identification method based on PSO-ELM. This method could reduce the hidden layer nodes of traditional ELM and improve the accuracy of identification. In this paper, this method was compared with ELM, GA-ELM, BP, GA-BP and PSO-BP. A sample of catenary detection data of a power supply section in 2018 was selected. The results show that PSO-ELM is an efficient method for the fault identification of catenary.

This paper proposes a novel fault diagnosis method for rolling bearing based on hierarchical refined composite multiscale fluctuation-based dispersion entropy (HRCMFDE) and particle swarm optimization-based extreme learning machine (PSO-ELM). First, HRCMFDE is used to extract fault features in the vibration signal at different time scales. By introducing the hierarchical theory algorithm into the vibration signal decomposition process, the problem of missing high-frequency signals in the coarse-grained process is solved. Fluctuation-based dispersion entropy (FDE) has the characteristics of insensitivity to noise interference and high computational efficiency based on the consideration of nonlinear time series fluctuations, which makes the extracted feature vectors more effective in describing the fault information embedded in each frequency band of the vibration signal. Then, PSO is used to optimize the input weights and hidden layer neuron thresholds of the ELM model to improve the fault identification capability of the ELM classifier. Finally, the performance of the proposed rolling bearing fault diagnosis method is verified and analyzed by using the CWRU dataset and MFPT dataset as experimental cases, respectively. The results show that the proposed method has high identification accuracy for the fault diagnosis of rolling bearings with varying loads and has a good load migration effect.

In the research of battery management driven by computer intelligent algorithms, Extreme Learning Machine (ELM) are suitable for lithium battery health assessment due to their high efficiency, but their single model is sensitive to parameters and has limited generalization ability. Therefore, this study proposes an integrated ELM model based on quantum particle swarm algorithm and K-means clustering, which realizes high-efficiency health assessment and life prediction of multi-parameter fusion by globally optimizing the base model parameters and clustering mining the data distribution law. In the performance test experiments, the fault identification accuracy rate of the model proposed by the research institute reached 0.99. In the 100 battery cycle aging experiment prediction scenarios, the average prediction error of the remaining battery service life was as low as 1, which was only 33.3% of similar models. The study provides a high-precision algorithm solution for the intelligent management of power lithium batteries. Its technical path of integrating swarm intelligence and clustering optimization provides new ideas for the application of artificial intelligence in the field of energy system status assessment.

This paper presents a model predictive control (MPC) strategy for three-level neutral-point-clamped ($3 ~\mathrm{L}-\text{NPC}$) converters. Conventional finite-control-set MPC (FCS-MPC) exhibits significant dependence on load parameters, resulting in limited robustness. To address this issue, we propose a dualvector model predictive control algorithm incorporating particle swarm optimization (PSO). The PSO algorithm is employed to dynamically identify load parameters, thereby eliminating reliance on predefined system parameters, while the proposed improved dual-vector method enhances prediction accuracy. The effectiveness of the proposed method is validated through MATLAB/Simulink simulations and experimental studies.

Distributed electric heating loads, vital demand response resources with significant adjustment potential, are critical for accurately identifying electric heating load model parameters to explore their energy-saving and demand response capabilities. This study proposes a second-order equivalent thermal parameter(ETP) identification method using an Improved Tyrannosaurus Rex Optimization Algorithm(ITROA). Dynamic weight coefficient-based hunting and chasing formulas adjust the TROA's step size for global and local optimization across iterations to prevent local optima convergence. Additionally, a Lévy flight strategy enhances global search randomness, prevents local optima and improves convergence performance. To validate the algorithm's effectiveness, parameter identification results from Particle Swarm Optimization(PSO), Genetic Algorithm(GA), standard TROA and the ITROA are compared using identical electric heating load data. Results demonstrate that the improved TROA significantly improves identification accuracy, offering substantial reference and application value.

In order to solve the problem that the load of cement ball mill is difficult to judge, based on particle swarm optimization(PSO) algorithm, a stacked convolutional autoencoder(SCAE) is proposed to judge the load. Firstly, an infinite impulse response(IIR) digital band-pass filter is designed to filter the vibration signals from the mill's wall, extracting vibration signals within an appropriate frequency range. Subsequently, two convolutional autoencoders are established to capture deep features from the vibration signals. Finally, the encoders from these two convolutional autoencoders are stacked together, followed by the addition of a classification layer to form the stacked convolutional autoencoder. This model is then utilized to classify the load state. PSO is employed to optimize both the filtering frequency band and the hyperparameters of the neural network. The results demonstrate that the stacked convolutional autoencoder achieves high accuracy, with an accuracy rate exceeding 97.53% on the test set. This provides effective assistance in judging the load state of cement ball mills and improving cement grinding quality.

Real-time monitoring of electricity usage details through load monitoring techniques is a crucial aspect of smart power grid management and monitoring, allowing for the acquisition of information on the electricity usage of individual appliances for power users. Accurate detection of electricity load is essential for refined load management and monitoring of power supply quality, facilitating the improvement of power management at the user side and enhancing power operation efficiency. Non-intrusive load monitoring (NILM) techniques require only the analysis of total load data to achieve load monitoring of electricity usage details, and offer advantages such as low cost, easy implementation, high reliability, and user acceptance. However, with the increasing number of distributed new load devices on the user side and the diversification of device development, simple load recognition algorithms are insufficient to meet the identification needs of multiple devices and achieve high recognition accuracy. To address this issue, a non-intrusive load recognition (NILR) model that combines an adaptive particle swarm optimization algorithm (PSO) and convolutional neural network (CNN) has been proposed. In this model, pixelated images of different electrical V-I trajectories are used as inputs for the CNN, and the optimal network layer and convolutional kernel size are determined by the adaptive PSO optimization algorithm during the CNN training process. The proposed model has been validated on the public dataset PLAID, and experimental results demonstrate that it has achieved a overall recognition accuracy of 97.26% and F-1 score of 96.92%, significantly better than other comparison models. The proposed model effectively reduces the confusion between various devices, exhibiting good recognition and generalization capabilities.

With the acceleration of urbanization, underground distribution cables have been widely adopted due to their economic efficiency and minimal spatial footprint. However, early-stage arc faults caused by insulation defects exhibit transient and concealed characteristics, while traditional detection methods suffer from insufficient identification accuracy due to susceptibility to fault types, locations, and noise interference. To address this, this paper proposes an early-stage cable fault identification method based on the fusion of a Particle Swarm Optimization (PSO) algorithm and a Deep Belief Network (DBN). The PSO optimizes the number of hidden-layer neurons in the DBN, and the model performance is enhanced through layer-wise pre-training and fine-tuning strategies. Based on a modified IEEE 13-node hybrid distribution network model, 8,856 samples-including half-cycle faults, multi-cycle faults, and motor-starting scenarios-were simulated to evaluate identification performance under varying noise levels. Experimental results demonstrate that the PSO-DBN achieves a fault recognition accuracy of 98.87%, surpassing the conventional DBN by 8.24%. Moreover, it maintains over 90% accuracy under 2-10 dB noise interference, significantly outperforming Extreme Learning Machine (ELM) and Probabilistic Neural Network (PNN) methods. These findings validate its advantages in feature extraction, nonlinear data processing, and robustness.

The issue of power quality disturbances in modern power systems has become increasingly complex and severe, with multiple disturbances occurring simultaneously, leading to a decrease in the recognition accuracy of traditional algorithms. This paper proposes a composite power quality disturbance identification method based on the integration of improved Complementary Ensemble Empirical Mode Decomposition (CEEMDAN), Hilbert Transform (HT), and Extreme Learning Machine (ELM). Addressing the limitations of traditional signal processing techniques in handling nonlinear and non-stationary signals, this study first preprocesses the collected initial power quality signals using the improved CEEMDAN method to reduce modal aliasing and spurious components, thereby enabling a more precise decomposition of noisy signals into multiple Intrinsic Mode Functions (IMFs). Subsequently, the HT is utilized to conduct a thorough analysis of the reconstructed signals, extracting their time-amplitude information and instantaneous frequency characteristics. This feature information provides a rich data foundation for subsequent classification and identification. On this basis, an improved ELM is introduced as the classifier, leveraging its powerful nonlinear mapping capabilities and fast learning speed to perform pattern recognition on the extracted features, achieving accurate identification of composite power quality disturbances. To validate the effectiveness and practicality of the proposed method, a simulation experiment is designed. Upon examination, the approach introduced in this study retains a fault diagnosis accuracy exceeding 95%, even amidst significant noise disturbances. In contrast to conventional techniques, such as Convolutional Neural Network (CNN) and Support Vector Machine (SVM), this method achieves an accuracy enhancement of up to 5%. Following optimization via the Particle Swarm Optimization (PSO) algorithm, the model’s accuracy is boosted by 3.6%, showcasing its favorable adaptability.

Arc fault in the three-phase load circuit may cause fire, resulting in production interruption and even worse, it will cause casualties. In order to effectively detect the arc fault in the three-phase circuit, series arc fault experiments of three-phase motor load and frequency converter were carried out under different current conditions. Firstly, variational mode decomposition (VMD) was performed for each cycle of A-phase current, and then the VMD energy entropy and sample entropy were calculated. Secondly, the noise-dominated component was removed according to the permutation entropy, then the average value after first-order difference of the half-cycle reconstructed signal was obtained. An arc fault diagnosis model of extreme learning machine (ELM) optimized by sparrow search algorithm (SSA) was established. The feature vectors were divided into training group and test group to train the model and test its fault diagnosis accuracy. Compared with GA-ELM, PSO-ELM, support vector machine (SVM) and SSA-SVM, the experimental results show that the proposed method can identify the series arc fault accurately and more quickly.

As a common load-bearing component, mining wire rope produces different types of damage during a long period of operation, especially in the case of damage inside the wire rope, which cannot be identified by the naked eye, and it is difficult to accurately detect such damage using the present technology. In this study we designed a non-destructive testing device based on leakage magnetism, which can effectively detect the internal defects of wire rope damage, and carried out simulation analysis to lay a theoretical foundation for the subsequent experiments. To address the noise reduction problem in the design process, a variational mode decomposition–adaptive wavelet thresholding noise reduction method is proposed, which can improve the signal-to-noise ratio and also calculate the wavelet energy entropy in the reconstructed signal to construct multi-dimensional feature vectors. For the quantitative identification of system damage, a particle swarm optimization–support vector machine algorithm is proposed. Moreover, based on the signal following the noise reduction step, seven different feature vectors, namely, the waveform area, peak value, peak-valley value, wavelet energy entropy classification, and identification of internal and external damage defects, have been determined. The results show that the device can be used to effectively identify internal damage defects. In addition, the comparative analysis showed that the algorithm can reduce the system noise and effectively identify internal and external damage defects with a certain superiority.

The transmitter and receiver for wireless power transfer (WPT) need information interaction to achieve optimal efficiency if the mutual inductance and load are movable. This paper proposes a receiver parameter estimation system based on the digital-twin concept aiming for condition monitoring of WPT. The digital replica and physical experimental demonstrate an application for a series-series WPT system. A hybrid optimization algorithm based on Particle Swarm Optimization and Simulated Annealing (PSO-SA) is applied to estimate mutual inductance and load based on the voltage and current of the transmitting coil from both the digital twin and the physical prototype. Compared with the conventional methods, the digital twin is real-time, without auxiliary circuits and invasion. In a validation experiment, the estimated parameters fit well with the actual values, verifying the accuracy of the proposed approach in comprehensively describing the state of the receiver. The outcomes of this study serve as a critical step for achieving noncommunication, cost-effectiveness, and power control for WPT.

No abstract available

No abstract available

Aiming at the problem that the existing asynchronous motor parameter identification method only identifies the parameters of the motor itself, ignoring the parameters of the asynchronous motor load, this paper proposes a particle swarm optimization algorithm combined with spatial disturbance to realize the integrated parameter identification of the asynchronous motor, machine and pump. On the premise of identifying the parameters of the asynchronous motor itself, the load parameters of the pump are identified, and the improved particle swarm optimization algorithm is used to realize the integrated identification of the asynchronous motor and the load. By combining the particle swarm optimization algorithm (PSO) and the spatial disturbance (SD), the six equivalent parameters of the asynchronous motor and the pump load factor can be accurately and effectively identified. Compared with the traditional PSO algorithm, the global search method is increased. Excellent ability. An example proves the effectiveness of the algorithm.

A high-voltage circuit breaker fault identification approach based on wavelet packet dispersion entropy feature extraction is suggested in order to address the issue of low accuracy of high-voltage circuit breaker fault diagnosis and identification under complicated operation states. Wavelet packet transform (WPT) is used first to break down the vibration signal. The feature matrix is then constructed by calculating the dispersion entropy of each subband. Lastly, the classification algorithm models of support vector machine (SVM), extreme learning machine (ELM), and particle swarm optimized extreme learning machine (PSO-ELM) are compared with the fault diagnosis model of particle swarm optimized support vector machine (PSO-SVM) to diagnose and classify each operating condition state of the circuit breaker. The PSO-SVM fault recognition model’s test set fault diagnostic accuracy is 95.83%, which is noticeably higher than that of the other three models, as demonstrated by the findings, confirming the method’s efficacy.

A Support Vector Machine (SVM) algorithm, combined with the Radial Basis Function (RBF), is employed in this paper for diagnosing power transformer oil through Dissolved Gas Analysis (DGA). The SVM algorithm's performance is enhanced by adjusting SVM parameter gamma “y” and the margin “C” for the RBF kernel. Such adjustments are achieved by coupling two optimization techniques; Particle Swarm Optimization (PSO) and Self-Learning Particle Swarm Optimizer (SLPSO). Compared to the conventional diagnosis techniques, improvements are obtained where an accuracy rate of 93.75% is achieved with the help of SLPSO optimizer.

Aiming at the fault diagnosis problem of oil-immersed transformers, this paper proposes a transformer oil dissolved gas analysis (DGA) fault diagnosis method based on the XGBoost algorithm. Traditional diagnostic methods have defects such as vague boundary values and poor interpretability. However, the XGBoost algorithm can effectively capture the nonlinear relationship between DGA data and fault types by iteratively optimizing the additive model. In the study, 1260 sets of DGA data were preprocessed, including mean normalization to eliminate the influence of dimensions, construction of gas ratio features to enhance sensitivity, and conversion of fault types into numerical labels. The model was optimized by setting hyperparameters such as learning_rate and max_ depth, and using 5-fold cross-validation and early stopping mechanism. Experimental results show that the accuracy of the XGBoost model on the test set reaches 93.6%, which is significantly higher than that of LSTM (82.3%) and PSO-LSTM (85.7%), and the RMSE and MAE indicators are better. The research shows that this method can accurately diagnose transformer faults and provide effective technical support for the safe operation of power systems.

In order to solve the problem that the redundancy of gas features in power transformer fault diagnosis will interfere with the parameter optimization process, and then affect the diagnosis accuracy, this paper proposes a fault diagnosis method based on feature optimization and BP neural network optimized by Seagull Optimization Algorithm (SOA) for dissolved gas analysis (DGA). Firstly, a 14-dimensional feature set was constructed by gas ratio method and sum normalization method. Secondly, the Random Forest (RF) algorithm is used to rank the importance of features and select $\mathbf{5 ~ k e y}$ features to reduce the interference of redundant information on the model. By comparing the diagnostic performance of support vector machine (SVM) and BP neural network, the results show that BP neural network has better classification ability after feature selection. Particle swarm optimization (PSO), Seagull optimization algorithm (SOA) and Condor optimization algorithm (BEO) were further introduced to optimize the parameters of BP neural network. The experiment showed that the diagnosis accuracy of SOA-BP neural network was the highest $(86.67 \%)$. Compared with the original BP neural network, PSO-BP and BEO-BP, the improvement is 10.25 %, 5.00 % and 3.34 %, respectively. The correctness of the method is verified by experiments.

Maintaining the safety and continuity of contemporary power systems depends critically on the accurate diagnosis of transformer failures. The most widely used diagnostic approach is still dissolved gas analysis (DGA); nevertheless, traditional ratio-based techniques, such as the Rogers’ ratio, rely on predefined thresholds and sometimes exhibit limited flexibility and unclear judgments under varied operating circumstances. This study suggests an optimization-oriented diagnostic approach that uses sophisticated metaheuristic algorithms to adaptively modify DGA gas ratio limitations in order to overcome these shortcomings. Four optimization schemes are formulated and comparatively assessed: the Artificial Protozoa Optimizer (APO), a hybrid Genetic Algorithm–Ant Colony Optimization model (GA–ACO), a hybrid Particle Swarm–Grey Wolf Optimization model (PSO–GWO), and a newly developed hybrid APO–PSO model. A dataset of 500 real-world DGA samples is used to evaluate the algorithms, and each optimization technique is conducted across 50 separate runs. The analysis focuses on statistical consistency, robustness, convergence characteristics, and diagnostic accuracy. With an average classification accuracy of around 96–97%, the suggested hybrid APO–PSO model outperforms standalone APO by about 2–3%, GA–ACO by 1–2%, and PSO–GWO by 1–2%, according to the numerical data. Furthermore, the APO–PSO scheme achieves more consistent behavior over repeated trials, reduced fitness variation, and quicker convergence. The statistical significance of these improvements is confirmed by statistical validation using the Friedman test and the Wilcoxon signed-rank test at a significance threshold of p < 0.05. Overall, the combination of APO’s strong global exploration with PSO’s efficient local exploitation produces a robust and adaptive diagnostic approach. The proposed framework enhances fault discrimination capability, reduces the likelihood of misclassification, and is suitable for both offline fault analysis and online transformer condition monitoring applications.

Dissolved gas analysis (DGA) provides valuable information for transformer condition monitoring, yet accurate multi-class fault identification remains challenging due to overlapping gas patterns and the sensitivity of classifier hyperparameters. This study proposes a hybrid optimization framework that combines Particle Swarm Optimization and Grey Wolf Optimization to tune the hyperparameters of a Support Vector Machine (SVM) for transformer fault diagnosis based on gas classification. The model is evaluated on a DGA dataset using a strict protocol that separates cross-validation–based tuning from held-out test assessment. Experimental results show that the proposed hybrid PSO-GWO-SVM achieves superior diagnostic performance and more stable convergence compared with representative single-optimizer baselines, demonstrating its potential for practical transformer fault identification.

In view of the shortcomings of traditional dissolved gas analysis technology low diagnostic veracity and low intelligence, this paper proposes to use QPSO to optimize the nuclear argument in the support vector machine (SVM), and on this basis, dissolved gas analysis (DGA) technology is used to diagnosis transformer faults. Firstly, the transformer data is preprocessed by DGA technology, and the processed data is used as the input amount of fault characteristics. Secondly, for the optimization of core parameters in SVM, the QPSO algorithm is combined with fault data for training and acquisition. Finally, five kinds of feature inputs are added to the model for training, and the trained multi-classification correlation vector machine is used to diagnose the test data. After case studies and comparative experimental analysis, the diagnostic accuracy of this method is as high as 94.74%, and relatively with SVM, PSO-SVM, and RVM methods, the accuracy is increased by 5.11%, 2.12%, and 2.12%, respectively.

To enhance the accuracy of transformer fault diagnosis, this study proposes an enhanced transformer fault diagnosis model incorporating the Improved Crow Search Algorithm (ICSA) and XGBoost. The dissolved gas analysis in oil (DGA) technique is employed to extract 9-dimensional fault features of transformers as model inputs, in conjunction with the codeless ratio method for training. The output layer utilizes a gradient boosting-based decision tree addition model to obtain the fault diagnosis type. Furthermore, the Golden Sine Algorithm (GSA) is employed for improvement, and the ICSA’s performance is tested by using typical test functions, demonstrating faster convergence and stronger merit-seeking capabilities. The obtained results reveal that the comprehensive diagnostic accuracy of the proposed model reaches 94.4056%, marking an improvement of 8.3916%, 6.2937%, 4.1958%, and 2.0979% compared to the original base XGBoost, PSO-XGBoost, GWO-XGBoost, and CSA-XGBoost fault diagnosis models, respectively. These findings validate the effectiveness of the proposed method in enhancing the fault diagnosis performance of transformers.

Aiming at the shortcomings of transformer fault diagnosis, a fusion algorithm (BWODO) based on Beluga algorithm (BWO) and Dandelion algorithm (DO) is proposed, which is combined with variational mode decomposition (VMD) and short and long time memory neural network (LSTM). Firstly, fault classification and feature extraction are carried out based on dissolved gas (DGA) data in oil. Then, the Dandelion algorithm is optimized through chaotic mapping opposition learning strategy, mixed reverse learning strategy, and Beluga whale predation strategy to improve the algorithm's optimization speed and solving accuracy. Then, the improved algorithm is used to optimize VMDLSTM model parameters to improve the model's accuracy of transformer fault identification. The test results show that the accuracy of BWODO-VMD-LSTM model is 97.4%, which is 5.9% and 4.2% higher than that of DO-VMD-LSTM and PSO-VMD-LSTM models, which proves that the proposed method can effectively improve the accuracy of transformer fault diagnosis.

This paper introduces a new methodology for diagnosing transformer failure to enhance diagnostic accuracy based on the combination of the Von Neumann Whale Optimization Algorithm (VNWOA) and the multi-categorical correlation vector machine (MRVM). Firstly, the whale algorithm is improved by using the principle of Von Neumann topology to increase the converge velocity and the optimization finding veracity of the whale algorithm by constructing a VN topology for each individual whale. Secondly, the VNWOA is applied to the MRVM arithmetic to find the optimal kernel features argument and penalty factor, and a diagnostic model of VNWOA-MRVM is advanced for failure diagnosis of the test data. Finally, the collected DGA data of 276 oil-immersed transformers are classified into exercise data and test data at a rate of 2: 1 for case analysis. After calculation, the accuracy of the proposed method can reach 95.65%, which is improved by 5.95%, 3.56%, 4.76% and 2.37% compared with SVM, PSO-SVM, M-RVM and PSO-RVM methods, respectively.

Machine learning based dissolved gas analysis (DGA) is a significant technique for the incipient fault diagnosis of power transformers. However, the diagnosis methods have limitations in learning from imbalanced fault dataset, especially in the case of a severe uneven distribution. The research interests, how to establish a suitable model for the imbalanced fault diagnosis, are drawing a great deal of attention. In this paper, a novel one-dimensional convolution neural network (1D CNN) model based on cost sensitive learning is proposed to be employed for transformer fault classification. Firstly, the structure of 1D CNN is designed with three hidden layers and two fully connected layers. Then, a class-dependent cost matrix is introduced into the soft-max function, which can modify the training process of the posed cost sensitive 1D CNN (CS-1D CNN), so as to pay more attention on the minority classes. Moreover, the PSO algorithm is adopted to optimize cost matrix for the CS-1D CNN. The performance of the proposed model is evaluated by case studies on a real-world fault dataset. The results reveal that the CS-1D CNN model has improved the accuracies of the minority fault classes and thus the overall accuracy of the whole dataset.

This paper proposes a hybrid Bat-BP approach based on dissolved gas-in-oil data set (DGA) to optimize the structure of back propagation neural network (BPNN). BPNN is a multilayer feed forward neural network. The rule of local decline that BPNN used is easy to fall into local optimum. Bat algorithm is a metaheuristic bionic algorithm with great local performance, which is adopted to optimize the initial value of BPNN. The recommended Bat-BP method has been employed in power transformer fault diagnosis for the first time. To prove the proposed method has better ability of power transformer fault diagnosis, this paper compares the fitness of Bat-BP with BPNN and other optimized approaches including PSO-BP, GA-BP based on the same DGA data set. The mean squared error (MSE) is used in this paper to evaluate the performance of the total four methods. The experimental results show the Bat-BP has increased the fault diagnosis accuracy from 75.68% to 95.22%, which is higher than those optimized models.

Early detection of power transformer fault is important because it can reduce the maintenance cost of the transformer and it can ensure continuous electricity supply in power systems. Dissolved Gas Analysis (DGA) technique is commonly used to identify oil-filled power transformer fault type but utilisation of artificial intelligence method with optimisation methods has shown convincing results. In this work, a hybrid support vector machine (SVM) with modified evolutionary particle swarm optimisation (EPSO) algorithm was proposed to determine the transformer fault type. The superiority of the modified PSO technique with SVM was evaluated by comparing the results with the actual fault diagnosis, unoptimised SVM and previous reported works. Data reduction was also applied using stepwise regression prior to the training process of SVM to reduce the training time. It was found that the proposed hybrid SVM-Modified EPSO (MEPSO)-Time Varying Acceleration Coefficient (TVAC) technique results in the highest correct identification percentage of faults in a power transformer compared to other PSO algorithms. Thus, the proposed technique can be one of the potential solutions to identify the transformer fault type based on DGA data on site.

Reliable bearing fault diagnosis is essential for the steady running of mechanical systems. However, existing diagnostic models still face significant limitations in feature extraction, primarily due to the non-stationary and nonlinear characteristics of vibration signals, which lead to a decline in diagnostic performance. To address this issue, this paper proposes a novel diagnostic framework that combines Particle Swarm Optimization-based Variational Mode Decomposition (PSO-VMD) for feature extraction with a deeply integrated Transformer-Convolutional Neural Network-Bidirectional Gated Recurrent Unit (TCB) model for fault classification. Bearing fault diagnosis is crucial for the stable operation of mechanical equipment, yet existing models often suffer from limited feature extraction and low detection accuracy. To address this, PSO-VMD is employed to extract informative, band-limited features from vibration signals, yielding a highly correlated feature set; a composite model TCB, combining a Transformer, a CNN, and a bidirectional GRU (BiGRU), is then used for fault classification. To prevent window-level leakage, the dataset is split before windowing and normalization, and all baselines are aligned under identical preprocessing and training settings. On the CWRU benchmark, the model attains 98.9% accuracy, 98.8% precision, 99.4% recall, 99.1% F1, and macro-F1 = 0.9766 over five runs. The approach offers a favorable accuracy –latency trade-off and yields interpretable, band-limited modes, supporting reproducible deployment in practice.

This paper introduces an intelligent fault-diagnosis framework for power transformers that integrates hybrid machine-learning models with nature-inspired optimization. Current signals were acquired from a laboratory-scale three-phase transformer under both healthy and various fault conditions. A suite of 41 discriminative features was engineered from time–frequency and sparse representations generated via Discrete Wavelet Transform (DWT) and Matching Pursuit (MP). The resulting dataset of 2400 labeled segments was used to develop four hybrid models, PSO-SVM, PSO-RF, BA-SVM, and BA-RF, wherein Particle Swarm Optimization (PSO) and the Bees Algorithm (BA) served as wrapper optimizers for simultaneous feature selection and hyperparameter tuning. Rigorous evaluation with 5-fold and 10-fold cross-validation demonstrated the superior performance of Random Forest-based models, with the BA-RF hybrid achieving peak performance (98.33% accuracy, 99.09% precision). The results validate the proposed methodology, establishing that the fusion of wavelet- and MP-based feature extraction with metaheuristic optimization constitutes a robust and accurate paradigm for transformer fault diagnosis.

Dissolved Gas Analysis (DGA) represents a widely recognized and routinely employed method for assessing the condition of oil-filled power transformers, particularly in the identification and classification of faults. In this research, a novel predictive framework is introduced for anticipating transformer anomalies, leveraging an advanced naïve Bayes classifier trained on DGA-derived datasets. The framework’s diagnostic accuracy and computational efficiency were further enhanced through the incorporation of multiple feature sets, whose dimensionality was systematically refined using a chaos-augmented binary Particle Swarm Optimization approach. This optimization procedure improved the exploration of the global search space, reduced the risk of premature convergence, and enabled the selection of highly informative features while simultaneously minimizing computational overhead. Additionally, a stringent cross-validation scheme was employed to ensure equitable segregation of training and testing data, thereby strengthening the reliability of the performance evaluation. Experimental results show that the proposed methodology consistently outperforms conventional approaches, achieving over 93% accuracy and delivering enhanced reliability in transformer fault diagnostics.

No abstract available

Transformer is a very important equipment in power system, its failure may lead to serious power system interruption and loss. Therefore, it is very important to develop an efficient and accurate transformer fault automatic diagnosis technology to ensure the reliable operation of the power system. This paper proposes a transformer fault automatic diagnosis technology based on particle swarm optimization (PSO) and extreme learning machine (ELM). By optimizing feature selection and using ELM algorithm for fault classification and diagnosis, the automatic diagnosis of transformer faults is made more rapid and accurate. The experimental data shows that the automatic diagnosis technology based on PSO-ELM has 99.6% accuracy and is 28.2% faster than the traditional technology, which provides strong support for the reliable operation of the power system. Future research can further explore the application potential of this method in other power equipment fault diagnosis and further improve its performance and reliability.

No abstract available

No abstract available

To overcome the shortcomings of the echo state network (ESN) algorithm in the application of fault diagnosis, a method of optimizing the ESN neural network, the particle swarm optimization (PSO) algorithm is established, and the transformer fault diagnosis algorithm is based on PSO-ESN. Through the simulation experiment of 50 groups of training samples and 15 groups of test samples, it can be seen that the PSO-ESN algorithm can improve the accuracy of transformer fault diagnosis and effectively reduce the error accuracy of the network.

In this paper a transformer fault diagnosis system based on a nature-based algorithm optimizing Support Vector Machine and Fuzzy Logic Model is proposed. Fault analysis and diagnosis are an integral part of operational reliability. Systems like SCADA collect data from various equipment in a power system network, however, cannot perform the critique fault diagnosis for the same. It, thereby, leads to additional costs. This paper uses the fuzzy model with a metaheuristic algorithm to build a predictive model for the data collected from various power transformers across Himachal Pradesh and IEC 10 database. A total of 225 datasets were collected and segregated into two sets. The datasets are created using the fuzzy model, IEC Ratio Method and concentration of key gases (ppm). Further, a support vector machine or SVM machine learning model is employed to classify the different faults in a transformer. The data is classified using binary and multiclass classification for an accurate diagnosis of transformer faults. The accuracy of SVM is improved by tuning its hyperparameters using Grid Search and Particle Swarm Optimization algorithm. A Classification Learner (MATLAB) model is also developed for the same dataset.

The optimized Radial Basis Function Neural Network (RBFNN) algorithm for Particle Swarm Optimization (PSO) is presented in this study in an effort to increase the precision of State of Charge (SoC) estimation in Lithium-Ion Batteries (LIBs). Using PSO, the RBFNN is known for its capability in modeling complex data relationships, is meticulously fine-tuned to significantly enhance the accuracy of SoC estimation. The process begins with the collection of raw sensor data, encompassing voltage, current, and temperature readings. This data undergoes a preprocessing phase designed to reduce noise and eliminate irrelevant information, ensuring that the RBFNN receives clean and relevant inputs. Following preprocessing, the RBFNN is subjected to rigorous training and testing phases. The integration of PSO optimizes the neural network's training process, leading to improved learning capabilities and more accurate SoC predictions. The model's performance is thoroughly evaluated against true SoC values using various statistical indices, including Root Mean Square Error (RMSE), Mean Absolute Error (AAE) and Average Absolute Error (AAE). Results indicate effectiveness of the PSO-RBFNN model in delivering precise SoC estimations, which are vital for optimizing management systems in battery and renewable energy applications. This research not only advances the field of battery management but also paves the way for future developments in energy storage technologies, highlighting the critical role of accurate SoC estimation in enhancing battery longevity and performance.

Lithium-ion batteries, being the main energy source for electric vehicles, need precise State of Charge (SOC) estimation to maintain system reliability. Traditional methods for SOC estimation do not meet the desired accuracy for engineering applications. This study introduces an advanced approach for SOC estimation in lithium-ion batteries by utilizing Particle Swarm Optimization (PSO) enhanced Particle Filtering. At the outset, a second-order RC equivalent circuit model is developed, and its parameters are determined. Considering significant differences in the open-circuit voltage (OCV) and SOC curves during charging and discharging, three types of OCV-SOC fitting curves are designed to define the state and observation equations. In conclusion, a particle filter-based SOC estimation model is built, and an improved PSO method is introduced to optimize state value determination. The experimental outcomes indicate that the proposed method surpasses conventional SOC estimation techniques in performance.

Accurate estimation of the State of Charge (SOC) for lithium-ion batteries is a core function of the Battery Management System (BMS). However, LiFePO4 batteries present specific challenges for SOC estimation due to the characteristic plateau in their open-circuit voltage (OCV) versus SOC relationship. Moreover, data-driven estimation approaches often face significant difficulties stemming from measurement noise and interference, the highly nonlinear internal dynamics of the battery, and the time-varying nature of key battery parameters. To address these issues, this paper proposes a Long Short-Term Memory (LSTM) model integrated with feature engineering, physical constraints, and the Extended Kalman Filter (EKF). First, the model’s temporal perception of the historical charge–discharge states of the battery is enhanced through the fusion of temporal voltage information. Second, a post-processing strategy based on physical laws is designed, utilizing the Particle Swarm Optimization (PSO) algorithm to search for optimal correction factors. Finally, the SOC obtained from the previous steps serves as the observation input to EKF filtering, enabling a probabilistically weighted fusion of the data-driven model output and the EKF to improve the model’s dynamic tracking performance. When applied to SOC estimation of LiFePO4 batteries under various operating conditions and temperatures ranging from 0 °C to 50 °C, the proposed model achieves average Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) as low as 0.46% and 0.56%, respectively. These results demonstrate the model’s excellent robustness, adaptability, and dynamic tracking capability. Additionally, the proposed approach only requires derived features from existing input data without the need for additional sensors, and the model exhibits low memory usage, showing considerable potential for practical BMS implementation. Furthermore, this study offers an effective technical pathway for state estimation under a “physical information–data-driven–filter fusion” framework, enabling accurate SOC estimation of lithium-ion batteries across multiple operating scenarios.

The maturity of lithium batteries has laid an important foundation for new energy and energy storage industries in recent years. Compared with other methods, lithium battery energy storage has the advantages of small volume, long circulation life, and high energy conversion rate. However, lithium batteries will affect the charging and discharge performance of lithium batteries at low temperature conditions, which will have a certain impact on the estimation of state of charge (SOC). Therefore, the estimation of lithium battery SOC at low temperature is meaningful. Home energy storage lithium batteries are different from vehicle power batteries and energy storage power station batteries. Its working environment is relatively stable and the temperature changes are not large, but there will still be low temperature conditions. This article uses a 100Ah lithium iron phosphate cell as a research object, and obtained the battery cell temperature, voltage, current, and internal resistance data of the battery during the low-temperature charging and discharge process. The back propagation (BP) neural network optimized by particle swarm optimization algorithm (PSO) is used to estimate the SOC of the battery. Compared with the general BP neural network and the long-short term memory networks (LSTM) neural network, the balance of the equity rectification is 42.55%and 95.97%.

The battery management system (BMS) is integral to the electric vehicle (EV) energy system, primarily responsible for managing the battery state and accurately estimating its state-of-charge (SOC). The precision of SOC estimation is critical for the accurate projection of the EV’s driving range and the optimal control of battery charging. To address the limited accuracy and inadequate adaptability of the existing SOC estimation algorithms, this article proposes a pioneering approach that combines the extended Kalman filter (EKF) algorithm with particle swarm optimization (PSO) and long short-term memory (LSTM) models to precisely estimate the SOC of power batteries. Validation demonstrates that the joint estimation algorithm maintains a root-mean-square error (RMSE) within 0.258% and a maximum error below 1.559% across various standard operating conditions and on-vehicle road testing (OVRT), signifying its excellent accuracy and robustness.

In order to achieve accurate state of charge (SOC) estimation of Lithium-Ion Battery, A method that dual Extended Kalman filters (DEKF) optimized by PSO-based Gray Wolf optimizer (MGWO) is proposed. A second-order equivalent circuit model with two resistor-capacitor branches is applied. The battery parameters are determined by battery test. Dual Extended Kalman filters are divided into state filter and parameter filter. Parameter filter is applied to adjust battery parameters online, state filter is applied to SOC estimation. Meanwhile, MGWO is applied to optimize the noise covariance matrix to improve the state estimation accuracy of SOC which reduces the linearization error from EKF. The results shows that the accuracy of algorithm is improved by adding online parameter identification and the optimization of the noise covariance matrix, meanwhile, the proposed method can adapt to the initial error well.

Accurately estimating the State of Charge (SOC) is of great significance for ensuring the safe operation of lithium-ion batteries and preventing overcharging or discharging. However, as SOC is the internal state of the battery unit and cannot be directly measured, it is difficult to obtain accurate SOC values. To improve the estimation accuracy of SOC, this paper establishes a prediction model using Kernel Extreme Learning Machine (KELM), and uses Particle Swarm Optimization (PSO) to optimize the kernel function parameter S and regularization coefficient C of KELM and determine the optimal values, a SOC prediction method for lithium-ion batteries based on PSO-KELM is proposed. The input of the PSO-KELM model is battery voltage, current, and temperature, the output is the actual SOC value. The results show that the root mean square error (RMSE), mean absolute error (MAE), and mean percentage error (MAPE) of SOC prediction are reduced to 2.76%, 2.18%, and 6.89%, increase coefficient of determination (R2) to 0.9919. The PSO-KELM model improves prediction accuracy compared to ELM and KELM models, and has good convergence and generalization properties.

In electric vehicles, accurate SOC can provide accurate and reliable residual power for drivers, which is the target value of a judgment method to effectively avoid overcharge or over discharge of battery pack[4]. However, due to the nonlinear characteristic of SOC value, its specific value cannot be directly measured by instruments and equipment. The final estimation of SOC value can only be carried out by monitoring battery parameters such as current and voltage. In order to achieve high-precision prediction of the state of charge (SOC) of lithium iron phosphate (LiFePO4) batteries, this paper uses an algorithm combining particle swarm optimization optimized by simulated annealing and BP neural network. Through MATLAB simulation, it can be seen that this method is more accurate than particle swarm optimization used to optimize BP neural network.

Lithium-ion battery State of Charge (SoC) estimate accuracy is importance for the effective and secure operation of Electric Vehicles (EVs), since it affects battery management and longevity. This study offers a revolutionary SoC estimation methodology that uses an Artificial Neural Network (ANN) optimized with Particle Swarm Optimization (PSO). The preprocessing includes resampling, outlier detection, and normalization, ensuring reliable input data. The ANN effectively captures the nonlinear dynamics of battery charge behavior, while PSO fine tunes the network’s hyperparameter to enhance prediction accuracy. The model implemented in Python, offers a reliable and efficient solution for real-world battery management applications, achieving a prediction accuracy of $94 \%$ and outperforming traditional estimation methods, thereby contributing to the advancement of reliable and intelligent battery monitoring systems in EV.

The accuracy of state-of-charge (SOC) estimation will affect the performance of the battery management system. The higher the accuracy the better the performance. To improve the accuracy of SOC estimation, a particle swarm optimization (PSO) based method is proposed to optimize the long short term memory. First, a PSO-Long Short Term Memory (LSTM) estimation model is established by the PSO algorithm, thereby achieving optimal iteration parameters of the model. Then, the PSO-LSTM estimation model is simulated under different working conditions and temperatures. Finally, the voltage, current, and other discharge data of the lithium-ion battery are input into the PSO-LSTM neural network model to compare with the LSTM algorithm. The results show that the estimation accuracy of the optimized PSO-LSTM algorithm model and extended Kalman filter is 2.1% and 1.5%, respectively. The accuracy is improved.

With the increasingly serious problem of environmental pollution, new energy vehicles have become a hot spot in today’s research. The lithium-ion battery has become the mainstream power battery of new energy vehicles as it has the advantages of long service life, high-rated voltage, low self-discharge rate, etc. The battery management system is the key part that ensures the efficient and safe operation of the vehicle as well as the long life of the power battery. The accurate estimation of the power battery state directly affects the whole vehicle’s performance. As a result, this paper established a lithium-ion battery charge state estimation model based on BP, PSO-BP and LSTM neural networks, which tried to combine the PSO algorithm with the LSTM algorithm. The particle swarm algorithm was utilized to obtain the optimal parameters of the model in the process of repetitive iteration so as to establish the PSO-LSTM prediction model. The superiority of the LSTM neural network model in SOC estimation was demonstrated by comparing the estimation accuracies of BP, PSO-BP and LSTM neural networks. The comparative analysis under constant flow conditions in the laboratory showed that the PSO-LSTM neural network predicts SOC more accurately than BP, PSO-BP and LSTM neural networks. The comparative analysis under DST and US06 operating conditions showed that the PSO-LSTM neural network has a greater prediction accuracy for SOC than the LSTM neural network.

Accurate state of charge (SOC) estimation is critical for the efficient management of lithium-ion batteries in electric vehicles and energy storage systems. This study proposes an improved SOC estimation framework by integrating the Extended Kalman Filter (EKF) with fuzzy logic control, optimized using Particle Swarm Optimization (PSO). The proposed method enhances estimation accuracy by dynamically adjusting the Kalman gain and optimizing model parameters. Simulation results demonstrate that the approach significantly reduces estimation errors and improves robustness under varying operating conditions. Compared to conventional methods, the proposed framework exhibits better adaptability to battery nonlinearities and environmental fluctuations. Despite its computational complexity, the method shows strong potential for practical applications in battery management systems. Future work will focus on optimizing real-time implementation and exploring machine learning-based enhancements to further improve performance.

Lithium-ion batteries are a possibility for alternative fuel engines. Accurate estimation of battery state of charge (SOC) ensures battery performance. This paper proposes a compression factor particle swarm optimization particle filter (FPSO-PF) algorithm for SOC estimation of lithium-ion batteries. To address the problems of particle swarm optimization particle filtering (PSO-PF), which is prone to fall into local optimum and unstable search speed, FPSO-PF improves the convergence of the algorithm by using a constraint factor and balances the global and local search capabilities. Finally, the FPSO-PF reduces the root mean square error (RMSE) and mean absolute error (MAE) of SOC estimation by 0.16 and 0.15%, respectively, compared with PSO-PF, as verified by experiments under dynamic stress test (DST) condition.

As an airborne emergency power supply, aviation lithium battery is the last line of defense for flight safety. In this paper, an online SOC estimation system was built based on improved BP neural network, and the initial weight and threshold of BP neural network were selected by GA-PSO algorithm. In addition, the acquisition of internal resistance parameters in the system was completed by the improved synchronous-sampling integration circuit. Through the experimental verification, the SOC online estimation system designed in this paper has high accuracy and reliability.

Accurate estimation of the state of charge (SOC) of lithium‐ion batteries is quite crucial to battery safety monitoring and efficient use of energy; to improve the accuracy of lithium‐ion battery SOC estimation under complicated working conditions, the research object of this study is the ternary lithium‐ion battery; the forgetting factor recursive least square (FFRLS) method optimized by particle swarm optimization (PSO) and adaptive H‐infinity filter (HIF) algorithm are adopted to estimate battery SOC. The PSO algorithm is improved with dynamic inertia weight to optimize the forgetting factor to solve the contradiction between FFRLS convergence speed and anti‐noise ability. The noise covariance matrixes of the HIF are improved to realize adaptive correction function and improve the accuracy of SOC estimation. To verify the rationality of the joint algorithm, a second‐order Thevenin model is established to estimate the SOC under three complex operating conditions. The experimental results show that the absolute value of the maximum estimation error of the improved algorithm under the three working conditions is 0.0192, 0.0131, and 0.0111, respectively, which proves that the improved algorithm has high accuracy and offers a theoretical basis for the safe and efficient operation of the battery management system.

State of charge (SOC) is an important parameter in battery management system, which is extremely important for evaluating the performance of lithium batteries. For the obvious drawbacks of integer-order lithium batteries and the bias of UKF in estimating SOC, this paper proposes an improved UKF algorithm based on the fractional-order model and uses the particle swarm algorithm (PSO) to improve the number of particles identified by the fractional-order leaping model for parameter identification. Comparison of the existing methods 2RC-UKF, 2RC-improved UKF, fractional-order UKF, and fractional-order-improved UKF shows that the average absolute error of fractional-order-improved UKF is 0.33%, and the RMSE is 0.21%, which indicates that the algorithm has excellent accuracy and robustness.

SOC reflects the residual storage volume of the battery and is an essential benchmark for preventing overcharging of the battery and predicting the remaining range of the vehicle. Thus, the accurate SOC estimation is critical for batteries. This paper proposes a methodology to predict SOC using FSSA to optimize the LSTM, and this method not only takes advantage of the LSTM in time series prediction, but also adopt the FSSA algorithm wisely which has strong search capability. Data under FUDS conditions are selected for training, and the performance of the different models are compared according to different evaluation indicators. The experimental data demonstrates that the predictive precision of this model is superior to the LSTM, PSO-LSTM and SSA-LSTM models, which demonstrates its robustness. Meanwhile, the experiments show that the FSSA-LSTM has 46.23% less computation time, and 40.73% less in MSE.

The state of charge (SOC) of lithium-ion batteries (LIBs) is a pivotal metric within the battery management system (BMS) of electric vehicles (EVs). An accurate SOC is crucial to ensuring both the safety and the operational efficiency of a battery. The unscented Kalman filter (UKF) is a classic and commonly used method among the various SOC estimation algorithms. However, an unscented transform (UT) utilized in the algorithm struggles to completely simulate the probability density function of actual data. Additionally, inaccuracies in the identification of battery model parameters can lead to performance degradation or even the divergence of the algorithm in SOC estimation. To address these challenges, this study introduces a combined UKF-LSTM algorithm that integrates a long short-term memory (LSTM) network with the UKF for the precise SOC estimation of LIBs. Firstly, the particle swarm optimization (PSO) algorithm was utilized to accurately identify the parameters of the battery model. Secondly, feature parameters that exhibited a high correlation with the estimation error of the UKF were selected to train an LSTM network, which was then combined with the UKF to establish the joint algorithm. Lastly, the effectiveness of the UKF-LSTM was confirmed under various conditions. The outcomes demonstrate that the average absolute error (MAE) and the root mean square error (RMSE) for the SOC estimation by the algorithm were less than 0.7%, indicating remarkable estimation accuracy and robustness.

Lithium-ion batteries (LIBs) are vital components in electric vehicles (EVs) and battery energy storage systems (BESS). Accurate estimation of the state of charge (SOC) and state of health (SOH) depends heavily on precise battery modeling. This paper examines six commonly used equivalent circuit models (ECMs) by deriving their impedance transfer functions and comparing them with measured electrochemical impedance spectroscopy (EIS) data. The particle swarm optimization (PSO) algorithm is first utilized to identify the ECM with the best EIS fit. Then, thirteen bio-inspired optimization algorithms (BIOAs) are employed for parameter identification and comparison. Results show that the fractional-order R(RQ)(RQ) model with a mean absolute percentage error (MAPE) of 10.797% achieves the lowest total model fitting error and possesses the highest matching accuracy. In model parameter identification using BIOAs, the marine predators algorithm (MPA) reaches the lowest estimated MAPE of 10.694%, surpassing other algorithms in this study. The Friedman ranking test further confirms MPA as the most effective method. When combined with an Internet-of-Things-based online battery monitoring system, the proposed approach provides a low-cost, high-precision platform for rapid modeling and parameter identification, supporting advanced SOC and SOH estimation technologies.

The state of charge (SOC) of lithium-ion batteries is a crucial parameter in battery management systems (BMSs). The particle swarm optimization (PSO) algorithm boasts advantages such as fast iteration speed and low computational complexity. However, a notable drawback of PSO is its tendency toward premature convergence. Therefore, this article presents a novel improved PSO. First, Lévy flight is used to generate random particles and exhibit wandering characteristics, thus effectively enhancing population diversity and avoiding the trapping of PSO in local optima. Second, the integration of dual-chaos theory with the golden sine algorithm (Golden-SA) optimizes the search performance of the particle swarm, with separate reconstructions of the fitness function and optimizations of the velocity update. Third, the bias-correction exponentially weighted moving average (BEWMA) method is further introduced to reduce the impact of noise and errors. It assigns reasonable weights to observation data at different time instances, enabling effective monitoring and propagation of varying errors. Ultimately, when data is used at $0~^{\circ }$ C, the root mean square error (RMSE) is 0.6844% and 0.4385%, respectively. The experimental results provide compelling evidence that they meet the operational requirements of BMSs.

Lithium‐ion batteries (LIBs) are the main energy source for electric vehicles (EVs), but they require sophisticated Battery Management Systems (BMS) for optimal functionality. In response to this need, the Python Battery Mathematical Model (PyBaMM) was used to apply the Doyle–Fuller–Newman (DFN) electrochemical model, which provided detailed battery data. This research utilizes the electrochemical DFN model to develop a surrogate model based on machine learning for precise state‐of‐charge (SoC) with predicted values of 15% to 90%, which is the recommended value of SoC in electric vehicle technology. The surrogate model showed impressive accuracy, achieving a 99.6% R‐score and a mean squared error (MSE) of 2.6%. Additionally, the study implemented a machine learning strategy integrated with particle swarm optimization (PSO) to determine optimal charging parameters that reduce charging time while preserving battery health and safety. These optimized parameters decreased the projected charging time to 130 s, although actual charging is expected to take around 225 s.

To address the problem of low State Of Charge (SOC) estimation accuracy caused by the strong nonlinearity of the lithium battery in electric vehicles. Compared with the estimation methods such as Back Propagation (BP), this paper proposes to take the lithium battery of electric vehicle as the research object, and adopt Particle Swarm Optimisation (PSO) to continuously optimise the weights and thresholds in the BP neural network, so as to improve its algorithm is easy to enter the situation of local extreme small value. Adaptive Inertia Weight Optimisation (AIW) is also used to improve the convergence and robustness of the particle swarm optimisation back propagation (PSO-BP)algorithm in the process of power battery SOC estimation, to prevent the algorithm from falling into an oscillating state. The interval cross-current charging and discharging data collected from electric vehicle lithium battery charging and discharging stations are used to estimate the SOC of electric vehicle power batteries in real time by learning the laws of historical data using the adaptive inertia weighted particle swarm back-propagation neural network fusion algorithm (AIW-PSO-BP). And the simulation is verified in Matlab/Simulink software. The simulation results show that based on the AIW-PSO-BP algorithm can well estimate the vehicle power battery SOC, with good convergence effect, and its absolute error is controlled at 0.4%, which effectively improves the prediction accuracy, and allows the vehicle to have a better method of estimating the SOC state.

Lithium-ion battery refers to a complex nonlinear system. Real-time diagnosis and accurate prediction of battery state of charge(SOC) parameters are hotspots and critical issues in battery research. To reduce the dependence of state of charge prediction on battery model accuracy and speed, and achieve real-time online estimation, a SOC prediction model of lithium-ion battery system is developed based on the model of support vector machine (SVM). SVM parameter is optimized using an algorithm of particle swarm optimization, and the performance of prediction model is assessed using cross-validation. The obtained experimental data is simulated, involving the comparison with the support vector machine model, and the prediction simulation of the battery in the state of fault. The results reveal that this model with a better performance than that of the support vector machine exhibits high accuracy and generalization ability.

Lithium-ion battery management is crucial as their use grows in devices and electric vehicles. A key aspect is State of Charge (SoC) estimation, which indicates the battery's charge level at any given time. This research aims to develop a method that can provide accurate SoC estimates for Li-ion batteries using the Particle Swarm Optimization (PSO) method. In this research, a 12V 8.4 Ah Lithium-Ion battery was used as a test subject, utilizing a voltage sensor, ACS712 sensor, and LM35 temperature sensor to measure key parameters such as voltage, current, and temperature. The PSO approach was chosen because of its ability to find optimal solutions in complex search spaces, such as SoC estimation in batteries. Through a combination of the PSO algorithm and data generated from sensors, it is hoped that the SoC estimates produced can improve battery usage efficiency, extend service life, and increase the performance of systems that depend on batteries. PSO can provide more accurate predictions with smaller errors, both in terms of the RMSE value of 0.0391 and the MAPE value of 12.028%. The high accuracy of 87.972% of PSO also shows that this method is reliable for applications that require precise SoC predictions. It is hoped that the results of this research can become a basis for further research in the field of battery management and metaheuristic algorithm optimization. After all, this research aims to enhance battery management systems and deepen understanding of PSO-based SoC estimation.

No abstract available

The state of charge (SOC) of lithium batteries is one of the important performance parameters of electric vehicles, and accurate real-time estimation of SOC can ensure the safe operation of electric vehicles. The traditional particle swarm optimization support vector machine algorithm is effective in predicting small samples. However, as the number of samples increases, there are problems in the prediction of lithium battery SOC of abnormal divergence in the later stage and unstable overall estimation results. To solve the above problems, this paper proposes a support vector machine model based on the natural selection method to improve the particle swarm optimization algorithm to realize the state-of-charge prediction of lithium batteries. The results of the simulation and test demonstrate that the method proposed in this paper can reduce the average relative error of prediction from 2.4% to 1.38%. The algorithm can improve the reliability and stability of the estimation results, and effectively guarantee the safe operation of electric vehicles.

Lithium-ion batteries are widely used in electronics, electric vehicles (EVs), and energy storage systems (ESS), but their performance declines over time, reflected in changes in state of health (SOH) and state of charge (SOC). Accurate estimation of these states is essential for reliable operation. This study evaluates several optimization algorithms for artificial neural networks (ANNs) and finds that particle swarm optimization (PSO) achieves the best performance, with mean relative errors of $\mathbf{0. 8 9 3 \%}$ for SOH and $\mathbf{1. 8 6 \%}$ for SOC, confirming its effectiveness for battery state estimation.

Addressing the issue that the estimation accuracy of underwater cabin-mounted lithium-ion battery status decreases as battery aging intensifies, a state estimation method based on a two-dimensional characterization factor of battery capacity increment is proposed. An equivalent circuit model of the battery is established, and the dependency of state of charge (SOC) estimation on the OCV-SOC curve is analyzed through theoretical derivation. A battery charging and discharging cycle test bench is set up to reveal the changing characteristics of the OCV-SOC curve during aging through experiments. The dQ/dV value of the highest peak and the sum of the voltage values corresponding to the two peaks in the battery capacity increment curve are selected as two characterization factors. A Gaussian process regression algorithm (GPR) based on particle swarm optimization (PSO) is designed. The algorithm is validated through two sets of charge-discharge data covering the entire life cycle of experimental batteries. The results indicate that as battery aging intensifies, the OCV-SOC curve undergoes significant and regular shifts. The extracted two-dimensional characterization factors exhibit a high correlation with the state of health (SOH). The designed GPR_PSO algorithm can accurately estimate the battery state and serves as an effective method for estimating the state of underwater cabin batteries.

The accurate state of charge (SOC) estimation for electric vehicle (EV) and superconducting magnetic energy storage (SMES) is quite important. For the SOC estimation with Kalman filter method, the inaccurate battery model results in estimation error. The traditional ampere-hour integral approach causes cumulative errors. The neural network method doesn't need precise battery model, but it needs much training time and low SOC estimating accuracy. The particle swarm optimization (PSO) algorithm for SOC estimation is proposed, since it can perform the nonlinear and dynamic characteristics of lithium-ion batteries, and the voltage, resistance and temperature are adopted as input vectors and the SOC is employed as output vector. The experimental results show that the estimation method for SOC in this paper has well precision with fast convergence.

Lithium-ion (Li-ion) batteries are the preferred choice for electric vehicles (EVs) because of their extended lifespans, low self-discharge rates, high voltage, and high energy density. A well-functioning Battery Management System (BMS) is critical to the efficient operation of an EV. The State of Charge (SoC) is an important statistic that reflects the remaining charge in the battery, and its exact assessment is essential for BMS and improving EV efficiency, which extends the battery's life and decreases the probability of catastrophic failure. However, SOC estimation is complicated and affected by numerous unknowns, such as battery age and external temperature. In this study, we estimated SOC using a Convolutional Neural Network (CNN) model. To improve the CNN architecture, this study has applied three different optimization algorithms: Particle Swarm Optimization (PSO), Elephant Search Algorithm (ESA), and Equilibrium Optimization (EO). Sensor data from lithium-ion batteries were carefully processed. The processed dataset was then supplied to the CNN and three optimized CNN models. These models were tested using error, R2, and time metrics to identify the optimal technique. CNN-ESA outperformed the other CNN models in SOC estimation, with the lowest error rates and the highest R2 value of 0.9987. This simulation result demonstrates the effectiveness of applying ESA to improve CNN architectures for better Li-ion battery SOC estimates. It enhances the efficiency and lifespan of EVs.

This paper proposes an improved method for estimating the state of charge (SOC) of lithium-ion battery. Firstly, a first-order resistor and capacitance (RC) model is introduced. Secondly, the SOC and open-circuit voltage (OCV) relationship is identified through the constant current charge-discharge test, and the least-squares algorithm is used to identify the model parameters. Thirdly, an improved adaptive approach is proposed to solve the problems of particle swarm optimization (PSO), and adaptive particle swarm optimized particle filtering (APSO-PF) is proposed to estimate the SOC of li-ion battery Finally, two dynamic operation conditions are given to show the efficiency of APSO-PF by comparing with the application of particle filter (PF), particle swarm optimized particle filtering (PSO-PF) and APSO-PF in SOC estimation.

Lithium-ion batteries have been widely used as energy storage systems and in electric vehicles due to their desirable balance of both energy and power densities as well as continual falling price. Accurate estimation of the state-of-charge (SOC) of a battery pack is important in managing the health and safety of battery packs. This paper proposes a compact radial basis function (RBF) neural model to estimate the state-of-charge (SOC) of lithium battery packs. Firstly, a suitable input set strongly correlated with the package SOC is identified from directly measured voltage, current, and temperature signals by a fast recursive algorithm (FRA). Secondly, a RBF neural model for battery pack SOC estimation is constructed using the FRA strategy to prune redundant hidden layer neurons. Then, the particle swarm optimization (PSO) algorithm is used to optimize the kernel parameters. Finally, a conventional RBF neural network model, an improved RBF neural model using the two stage method, and a least squares support vector machine (LSSVM) model are also used to estimate the battery SOC as a comparative study. Simulation results show that generalization error of SOC estimation using the novel RBF neural network model is less than half of that using other methods. Furthermore, the model training time is much less than the LSSVM method and the improved RBF neural model using the two-stage method.

Significant improvements in battery performance, cost reduction, and energy density have been made since the advancements of lithium-ion batteries. These advancements have accelerated the development of electric vehicles (EVs). The safety and effectiveness of EVs depend on accurate measurement and prediction of the state of health (SOH) of lithium-ion batteries; however, this process is uncertain. In this study, our primary goal is to enhance the accuracy of SOH estimation by reducing uncertainties in state of charge (SOC) estimation and measurements. To achieve this, we propose a novel method that utilizes the gradient-based optimizer (GBO) to evaluate the SOH of lithium batteries. The GBO minimizes a cost with the aim of selecting the optimal candidate for updating the SOH through a memory-fading forgetting factor. We evaluated our method against four robust algorithms, namely particle swarm optimization-least square support vector regression (PSO-LSSV), BCRLS-multiple weighted dual extended Kalman filtering (BCRLS-MWDEKF), Total least square (TLS), and approximate weighted total least squares (AWTLS) in hybrid electric vehicle (HEV) and electric vehicle (EV) applications. Our method consistently outperformed the alternatives, with the GBO achieving the lowest maximum error. In EV scenarios, GBO exhibited maximum errors ranging from 0.65% to 1.57% and mean errors ranging from 0.21% to 0.57%. Similarly, in HEV scenarios, GBO demonstrated maximum errors ranging from 0.81% to 3.21% and mean errors ranging from 0.39% to 1.03%. Furthermore, our method showcased superior predictive performance, with low values for mean squared error (MSE) (<1.8130e-04), root mean squared error (RMSE) (<1.35%), and mean absolute percentage error (MAPE) (<1.4).

The state of charge (SOC) is a critical parameter in the battery management system (BMS), and its accurate estimation is essential for ensuring the safety and reliability of batteries. This paper presents a lithium-ion battery SOC estimation method that combines an improved neural network with a filtering algorithm. Firstly, the backpropagation (BP) algorithm is chosen as the architecture of the neural network in the hybrid method due to its strong nonlinear approximation ability, and the particle swarm optimization (PSO) algorithm is used to optimize it to avoid falling into local optimal solutions. By combining the search ability of PSO with the learning ability of the BP neural network, the accuracy of the neural network model is improved. The proposed method integrates the PSO-BP model with the extended Kalman filter based on minimum error entropy (MEE-EKF). PSO-BP is utilized as the measurement equation for MEE-EKF, while the ampere-hour integration method is employed as the state equation to achieve closed-loop SOC estimation. Finally, experimental validation is conducted under four typical operating conditions and one random condition across a wide temperature range. The results demonstrate that the proposed method achieves high accuracy across all conditions compared with the results of other algorithms, with a maximum absolute error of not exceeding 3.13%, a mean absolute error of less than 0.54%, and a root mean square error of no more than 0.66%.

High‐precision state of charge (SOC) estimation is essential for battery management systems (BMSs). In this paper, a new SOC estimation method is proposed. The dual Kalman filter algorithm and backpropagation neural network (particle swarm optimization ‐ backpropagation neural network ‐ double extended Kalman filter [PSO‐BPNN‐DEKF]) are combined to estimate and correct the SOC of lithium‐ion batteries, in which the initial weight and threshold of the BPNN are optimized by particle swarm optimization algorithm. Based on the second‐order RC equivalent circuit model, parameter identification is carried out using the adaptive forgetting factor least squares (AFFRLS) method. Online parameter updates and SOC estimation are realized by DEKF algorithm. Then, the trained PSO‐BPNN is used to predict the SOC estimation error in real time, and the SOC estimation value is corrected by adding prediction errors. The SOC estimates before and after correction under Beijing Dynamic Stress Test (BBDST), dynamic stress test (DST), and hybrid pulse power characterization (HPPC) were compared. Under BBDST, DST, and HPPC tests, the maximum errors of the corrected SOC estimates are 0.0107, 0.0090, and 0.0147, respectively. The root mean square error (RMSE) of the corrected SOC estimates decreased by 94.02%, 83.18%, and 88.03%, respectively, compared with the extended Kalman filtering (EKF). The mean absolute error (MAE) of the corrected SOC estimates remained around 0.1% for all the BBDST dynamic operating conditions at different temperatures. The experimental results demonstrate the accuracy, effectiveness, and temperature adaptability of the proposed algorithm for SOC estimation under complex conditions of lithium‐ion batteries.

The precise assessment of the state of charge (SOC) of lithium-ion batteries (LIBs) is critical in battery management systems. This work offers a comprehensive learning particle swarm optimization (CLPSO) and extended Kalman filter (EKF) technique to forecast the SOC of LIBs in order to obtain an accurate SOC estimate for power batteries. First, to address the challenge of identifying various parameters of the battery model, the bilinear transformation technique is employed to determine the parameters of the second-order RC equivalent circuit model. Second, to improve the fitness values for the conventional PSO algorithm, which is prone to entering local optimality, a learning strategy (f_i) is added to the particle velocity update method. The optimized PSO and EKF algorithms are integrated to perform online prediction of the SOC of LIBs. The experimental results demonstrate that under the conditions of the Beijing Bus Dynamic Stress Test (BBDST), Dynamic Stress Test (DST), and Hybrid Pulse Power Characterization Test (HPPC), the parameter identification inaccuracy of CLPSO is restricted to 1%. After multi-metric evaluation, the maximum error and mean absolute error of the CLPSO-EKF algorithm in SOC estimation are 0.32% and 0.0652%, respectively, demonstrating a higher robustness and accuracy advantage over other versions.

User-side carbon emission accounting is a means to sort out the amount and source of user carbon emissions. Real-time carbon estimation methods based on nonintrusive load monitoring (NILM) tend to miss the fact that there is not a negative correlation between equipment identification accuracy and carbon emission estimation error. Higher accuracy does not mean lower error. A two-stage training NILM network, LRMSE-ResNeSt, is proposed to reduce the error of real-time carbon emission monitoring results. It emphasizes the low root-mean-square error (RMSE) of carbon emissions’ monitoring. The feature extraction network is first constructed using convolution and split-attention mechanism, and then, the model is trained using a two-stage training approach. In the first stage, backpropagation is used to update the network parameters for accurate device identification. In the second stage, the parameters of the fully connected layer are tuned using particle swarm optimization (PSO) to make the classifier more focused on the identification accuracy of devices with high carbon emissions. Finally, the proposed LRMSE-ResNeSt is validated using the industrial appliance identification dataset (IAID) industrial dataset. The experimental results show that the LRMSE-ResNeSt successfully reduces the RMSE of real-time carbon estimation by an average of 14.94%, which addresses the problem that the NILM method only focuses on the accuracy of the equipment identification but ignores the error of the carbon emission estimation results.

Smart energy management demands better ways to understand the energy consumption of buildings. Nonintrusive Load Monitoring (NILM) is an emerging technique that disaggregates total building consumption to individual appliances. However, low-power appliances pose a challenge as they exhibit similar power consumption patterns that are difficult to identify and distinguish from background noise. This research aims to bridge this gap by developing feature extraction and classification techniques specifically for low-power NILM. This may facilitate the development of targeted energy-saving strategies and result in more precise energy-use monitoring. To effectively identify and differentiate low-power appliances within a NILM system, this work proposes a novel feature extraction approach that combines mono-fractal and multifractal analysis of appliance startup current transients. Mono-fractal features, including fractal dimension and Hurst exponent, are extracted alongside Lacunarity and Multifractal features comprising the singularity spectrum and Hölder exponents. These features collectively create comprehensive appliance signatures that capture the unique characteristics of each appliance during its turn-on event. Building upon these extracted features, the work investigates the performance of three machine learning classifiers for appliance classification: deep neural network, support vector machine, and K-nearest neighbours. These classifiers are optimized using Bayesian optimization to achieve optimum performance. The proposed method demonstrates significant improvement over existing feature extraction approaches across all optimized classifiers, achieving an accuracy of up to 98.3% in classifying low-power appliances.

Non-intrusive load monitoring (NILM) and identification are important to smart grid construction and safe operation. A method combining mixed feature matrix and convolutional neural network based on V-I trajectory and odd harmonic components is proposed and proved for non-intrusive electrical load monitoring technology. In the paper, the specific contents are to analyze the basic characteristics of electrical equipment, including steady-state and transient characteristics of input current. Six simulation circuits of single-phase AC power supply powered electrical equipment are established by means of MATLAB/Simulink using Electrical or Electronic blockset, including heater, hair dryers LED string, adapter, variable frequency air conditioner (air-CON), and variable frequency refrigerators, as the electrical equipment to be analyzed and identified. The grid side current waveforms and spectrums, V-I trajectories, and odd harmonic current components of these six load models are extracted as basic features. The mixed feature matrix using V-I trajectory and odd harmonic features, as well as convolutional neural network (CNN), are proposed for the analysis and identification of electrical equipment. After training and testing on the PLAID dataset, the system can effectively identify electrical equipment under steady-state operation. In addition, three different detection methods are compared and analyzed to verify the superiority of the proposed method for analyzing and identifying electrical equipment based on mixed feature matrix and convolutional neural network.

Infrared thermal image technology is widely used for detecting the state of electrical equipment due to its noncontact advantages. However, as the scale of electrical equipment continues to expand, the analysis of infrared images, characterized by big data, still relies on manual methods, resulting in low efficiency and high costs. Therefore, this paper proposes an infrared image identification method for electrical equipment based on a deep convolution neural network. First, infrared images of electrical equipment, such as insulators, transformers, and circuit breakers, are collected, and dataset augmentation techniques like cropping and rotation are applied. A deep convolutional neural network model is then established by incorporating a multi-scale approach, with training conducted using cross-entropy loss functions. Finally, a coordinated attention mechanism based on regions of interest is introduced to enhance identification accuracy. The results demonstrate that the proposed method achieves a mean Average Precision (mAP) of $\mathbf{9 3. 6 4 \%}$ in validation, with the coordinated attention mechanism improving the mAP by $4.24 \%$. Thus, the proposed method significantly enhances identification accuracy and lays a foundation for the condition monitoring of electrical equipment.

A methodology for identifying the operating modes of a CNC metal-cutting lathe in production conditions has been formed. Methods of obtaining indirect information about the state of the production machine in order to analyze it at different levels of its depth are considered. The possibility of identifying the operating modes of a CNC lathe based on data on the change in electric current consumed by the main motion drive is justified. A hardware and software complex has been developed to collect information about changes in electrical energy in the machine power supply network for a long time. Experimental data have been collected on the change in electric current consumed by production equipment during its continuous operation for several weeks. Data processing in the Matlab application package was performed to assess the adequacy of the proposed identification technique. A numerical assessment of the optimal distribution of time resources of the metal cutting machine during the working day was performed. Prospects and opportunities for developing a methodology based on a deeper analysis of data on changes in electric current with the aim of subsequent construction of intelligent systems for monitoring and forecasting the state of production equipment and metalworking processes are considered.