天气因素对车辆出行影响的研究现状

ABSTRACT Winter hazards can lead to traffic safety issues, often resulting from uninformed travel decisions. To address these challenges, the Intelligent Winter Road Information System (IWRIS) was developed to improve driver awareness by providing supportive information and timely alert notifications through various media platforms. This study utilizes a stated preference survey to examine how travelers respond to IWRIS and assesses its impact on driver decision-making during winter conditions. The results indicate a strong correlation between the consistent acceptance of navigation system recommendations, overall system usage, and critical variables, such as weather conditions, driver experience, route familiarity, and vehicle characteristics. The study underscores the necessity of understanding the impact of information provision on travel choices in challenging winter conditions and highlights the need for systems that effectively adapt to the unique challenges of winter.

The increasing adoption of electric vehicles (EVs) presents significant implications for transportation patterns, particularly in route choice behavior. This research investigates the factors affecting EV drivers' route selection decisions for long-distance travel, focusing on charging infrastructure accessibility, environmental awareness, and range anxiety concerns. We conducted a combined revealed preference (RP) and stated preference (SP) survey using D-efficient experimental design to analyze route choice preferences among 600 EV drivers undertaking trips over 100km. The experimental design incorporates multiple attributes such as route type, charging station availability, traffic conditions, weather patterns, and battery state of charge. Our approach enables comprehensive examination of both infrastructure factors and environmental considerations in EV route selection behavior. Route choice data analysis employed multinomial logit modeling techniques. Findings reveal that travel cost has a strong negative influence on route selection, while charging infrastructure density demonstrates minimal impact on route choice decisions. Traffic flow patterns and weather conditions constitute key factors in route decision-making processes. The study makes three primary contributions to existing literature: first, through integrated analysis of infrastructure availability and environmental consciousness in EV route selection while considering vehiclespecific attributes; second, by demonstrating empirical relationships between charging infrastructure characteristics and EV route choice patterns; and third, through quantitative assessment of connections between environmental awareness and actual routing decisions among EV drivers. These results enhance theoretical understanding of EV route choice behaviors and offer evidence-based insights for strategic charging infrastructure development during the transition toward sustainable mobility systems.

Vehicle re-identification aims to match and identify the same vehicle crossing multiple surveillance cameras and obtain traffic information such as travel time. The Automatic License Plate Recognition (ALPR) data are widely employed in urban surveillance. However, vehicle re-identification based on ALPR data is challenging due to license plate recognition errors and unrecognized issues. This paper proposes a vehicle matching algorithm designed to maximize the travel time probability using ALPR data, while accounting for recognition errors and unrecognized issues. The proposed algorithm consists of several modules, including the estimation of travel time distribution, computation of travel time probability, calculation of travel time confidence intervals and matching time window size, restricted fuzzy matching, and vehicle matching optimization. To evaluate the effectiveness of the proposed algorithm across varying lighting and weather conditions, ALPR data was collected from a survey road in four scenarios: sunny day, sunny night, rainy day, and rainy night. The results indicate that when compared to a sunny day scenario, severe lighting and adverse weather conditions lead to decreased matching accuracy and increased matching accuracy errors for all methods evaluated. However, our proposed model outperforms benchmark algorithms in both scenarios, demonstrating its superior performance.

In this study, a spatio-temporal prediction model for EV charging loads is proposed based on the Graph diffusion convolution-Long short-term memory network (GDC-LSTM). This model integrates diverse sources of information, including weather conditions, day types, and temperatures. Firstly, the GDC-LSTM is employed to predict EV traffic flow data, capturing the travel patterns of EV users. Subsequently, a real-time model for EV power consumption is developed, and the floyd shortest path algorithm is optimized by introducing real-time road saturation data.. Finally, employing the Monte Carlo method, spatio-temporal distributions of various EV charging loads are forecasted and mapped onto distribution network nodes. The effectiveness of the methodology is verified by demonstrating the model's ability to predict EV travel behavior and load distribution through a case study in the main urban area of Chengdu, and evaluating the impact of load access on distribution network node voltages.

Investigating travel time variability is critical for pre-trip planning, reliable route selection, traffic management, and the development of control strategies to mitigate traffic congestion problems cost-effectively. Hence, a large number of studies are available in the literature which determine the most suitable distribution to fit the travel time data, but these studies recommend different distributions for the travel time data, and there is a disagreement on the best distribution option for fitting to the travel time data. The present study proposes a novel framework to determine the best distribution to represent the travel time data obtained from probe vehicles by using the modern machine learning technique. This study employs vast travel time data collected by fitting GPS tracking units on the probe vehicles and offers a comprehensive investigation of travel time distribution in different scenarios generated due to spatiotemporal variation of the travel time. The study also considers the effect of weather and uses the three most commonly used non-parametric goodness-of-fit tests (namely, Kolmogorov–Smirnov test, Anderson–Darling test, and chi-squared test) to fit and rank a comprehensive set of around 60 unimodal statistical distributions. The framework proposed in the study can determine the travel time distribution with 91% accuracy. Additionally, the distribution determined by the framework has an acceptance rate of 98.4%, which is better than the acceptance rates of the distributions recommended in existing studies. Because of its robustness and applicability in many different traffic situations, the proposed framework can also be used in developing countries with heterogeneous disordered traffic conditions to evaluate the road network’s performance in terms of travel time reliability.

With the development of intelligent transportation system, more and more traffic detectors have been deployed in urban road networks. However, the failure rate of equipment such as automatic license plate reader (ALPR) detector is high after long-term use. Besides, bad weather, license plate occlusion and other factors will affect the detection accuracy. These above reasons will lead to the discontinuity of vehicle trajectories, which will lead to huge errors in the calculation of traffic parameters such as traffic flow. Moreover, trajectory missing is not conducive to the supervision of traffic management departments. Therefore, this paper proposes a tensor based individual travel time matching method to recover missing vehicle trajectories. An initialized tensor will be used to estimate the travel time of every particular vehicle whose trajectory missed, and then the most likely trajectory will be calculated to recover the missing parts. Finally, the proposed method performs well in the ALPR dataset measured in Ruian, Zhejiang Province, China.

This study investigates the factors influencing school trip mode choice among senior high school students in the Chiang Rai urban area, Chiang Rai, Thailand, under normal and adverse weather conditions. Utilizing data from 472 students across six extra-large urban schools, a Multinomial Logit (MNL) regression model was applied to examine the effects of socio-demographic attributes, household vehicle ownership, travel distance, and spatial variables on mode selection. The results revealed notable modal shifts during adverse weather, with motorcycle usage decreasing and private vehicle reliance increasing, while school bus usage remained stable, highlighting its role as a resilient transport option. Car ownership emerged as a strong enabler of modal flexibility, whereas students with limited access to private transport demonstrated reduced adaptability. Additionally, increased waiting and travel times during adverse conditions underscored infrastructure and service vulnerabilities, particularly for mid-distance travelers. The findings suggest an urgent need for transport policies that promote inclusive and climate-resilient mobility systems, particularly in the context of Chiang Rai, including expanded school bus services, improved first-mile connectivity, and enhanced pedestrian infrastructure. This study contributes to the literature by addressing environmental variability in school travel behavior and offers actionable insights for sustainable transport planning in secondary cities and border regions.

In real-life logistics, distribution plans might be affected by weather conditions (rain, snow, and fog), since they might have a significant effect on traveling times and, therefore, on total distribution costs. In this paper, the distribution problem is modeled as a multi-depot vehicle routing problem with stochastic traveling times. These traveling times are not only stochastic in nature but the specific probability distribution used to model them depends on the particular weather conditions on the delivery day. In order to solve the aforementioned problem, a simheuristic approach combining simulation within a biased-randomized heuristic framework is proposed. As the computational experiments will show, our simulation-optimization algorithm is able to provide high-quality solutions to this NP-hard problem in short computing times even for large-scale instances. From a managerial perspective, such a tool can be very useful in practical applications since it helps to increase the efficiency of the logistics and transportation operations.

This study examined the effectiveness of weather warning messages posted on a changeable message sign (CMS) as a speed reduction countermeasure for motorists approaching freeway bridge overpasses during winter weather conditions. Two weather warning messages were tested, including “Bridge Ices Before Road / Reduce Speeds” and “Slippery Road Conditions / Reduce Speeds,” in addition to the default travel time messages. The messages were evaluated through a series of field studies conducted in Michigan at three rural freeway bridge overpasses with a CMS on the approach during the winter seasons of 2023 and 2024. Speed measurements were obtained for each subject vehicle both upstream of the CMS and also at the start of the bridge. The study found that the winter weather warning messages were effective toward reducing the speed of vehicles approaching the bridge overpass during winter weather conditions. The most substantial impacts, in relation to both the speed reduction likelihood and magnitude, were observed when the “Slippery Road Conditions / Reduce Speeds” message was displayed on the CMS. With this message displayed, drivers were twice as likely to reduce their speed before reaching the bridge and the speed reductions were 0.6 to 0.7 mph greater, on average, compared with the default travel time messages. In addition, the strongest speed reduction were observed for the fastest group of drivers, which is the most vulnerable group from a crash severity risk standpoint and, thus, are considered the most highly targeted driver group for speed reductions.

Recent advancements in object detection and deep learning have significantly enhanced intelligent transportation systems, contributing to safer and more efficient roadways. This study proposes a comprehensive framework that leverages YOLOv8, augmented with vehicle tracking and a dehazing module, to detect and track moving vehicles under various weather conditions. The system can accurately identify vehicle types such as cars, motorcycles, buses, or trucks, track their trajectories, and estimate their speeds. It includes alert mechanisms that notify when vehicles exceed or fall below speed limits or travel in the wrong direction. To ensure robustness in adverse weather, the framework incorporates a hybrid loss function with both pixel and structure similarity measurements to train the dehazing model. This dehazing model is capable of effectively mitigating the effects of haze, fog, and rain in video streams. This work highlights the potential of AI-driven solutions for real-time vehicle monitoring, risk mitigation, and the advancement of road safety in adverse weather.

Living in large cities is prone to adverse environmental impacts including air pollution. Increased vehicle traffic generates congestion as some road infrastructure such as intersections, including signalised ones, have limited capacity. When congestion exceeds capacity, vehicle queues occur and their presence increases delays and generates more harmful exhaust fumes. Travel costs also increase. Studies show that adverse weather conditions negatively affect drivers’ behaviour, who become more cautious and slower to exit intersections during the green signal. Such behaviour further reduces lane capacity, increasing congestion relative to similar level of demand flow on days with favourable weather (sunny and cloudy days, no rain). The paper will demonstrate the impact of weather conditions on lane capacity and traffic performance on signalised through lanes. Worse traffic performance on days with prolonged rainfall or snowfall contributes to increased travel costs and air pollution, as will also be demonstrated in the paper.

Synergetic transportation schemes are extensively used in package delivery operations, exploiting the best features of different modes. This paper proposes a methodology to solve the mode assignment and routing problem for the case of combined conventional vehicle and unmanned aerial vehicle (CV–UAV) parcel deliveries under uncertainty for next-day operations. This research incorporates ground and air uncertainties: travel times are assumed for conventional vehicles, while UAV paths are affected by weather conditions and restricted flying zones. A nested genetic algorithm is initially used to solve the problem under fixed conditions. Then, a robust optimization approach is employed to propose the best solution that will perform well in a stochastic environment. The framework is applied to a case study of realistic urban–suburban size, and results are discussed. The entire platform is useful for strategic decisions on infrastructure and for operation planning with satisfactory performance and less risk.

This study addresses the Capacitated Vehicle Routing Problem with Pickup and Delivery (CVRPPD), a core challenge in urban logistics involving the optimization of vehicle routes under dynamic constraints. Traditional algorithms predominantly focus on static variables like distance, failing to account for real-world factors such as traffic congestion, adverse weather, and vehicle capacity limitations. To solve this problem, the Adaptive Heuristic-Based Ant Colony Optimization (AHB-ACO) algorithm was developed, incorporating these dynamic constraints into the routing optimization process. The AHB-ACO algorithm minimizes total travel costs while ensuring adherence to vehicle capacity limits and improving route safety. Simulation tests were conducted on datasets with 50, 100, and 200 customers to evaluate performance under varying levels of complexity. The results demonstrate that AHB-ACO outperforms traditional ACO, particularly in dynamic scenarios, achieving a total cost of 4155.82 with an execution time of 1639.68 seconds for the 200-customer dataset. The algorithm’s adaptive heuristic formula integrates distance, traffic congestion, and weather penalties, enabling the generation of safer and more realistic routes. These results are explained by the algorithm’s ability to dynamically adjust to constraints, ensuring robust performance in complex environments. The findings highlight AHB-ACO’s practical applicability in urban logistics, offering scalability and adaptability for real-world delivery and pickup challenges, especially in areas affected by fluctuating traffic and weather conditions

Adaptive Cruise Control (ACC) can change the speed of the ego vehicle to maintain a safe distance from the following vehicle automatically. The primary purpose of this research is to use cutting-edge computing approaches to locate and track vehicles in real time under various conditions to achieve a safe ACC. The paper examines the extension of ACC employing depth cameras and radar sensors within Autonomous Vehicles (AVs) to respond in real-time by changing weather conditions using the Car Learning to Act (CARLA) simulation platform at noon. The ego vehicle controller's decision to accelerate or decelerate depends on the speed of the leading (ahead) vehicle and the safe distance from that vehicle. Simulation results show that a Proportional–Integral–Derivative (PID) control of autonomous vehicles using a depth camera and radar sensors reduces the speed of the leading vehicle and the ego vehicle when it rains. In addition, longer travel time was observed for both vehicles in rainy conditions than in dry conditions. Also, PID control prevents the leading vehicle from rear collisions.

Real-world Vehicle Routing Problems (VRPs) often require computing travel times between tens of thousands of geographic locations to plan routes for hundreds of vehicles. For example, during the COVID-19 vaccine distribution in Korea, hundreds of vehicles were tasked with delivering vaccines to thousands of destinations each week. Solving such VRPs necessitates estimating pairwise travel times between origin-destination (OD) pairs, a process that scales quadratically with the number of destinations. Existing travel time estimation methods typically focus on individual routes and rely on rich contextual data such as GPS trajectories, road types, traffic volume, and weather conditions. While accurate, these methods are computationally expensive and impractical for large-scale VRPs due to their input requirements and inefficiency. To address these challenges, this paper proposes an efficient and scalable deep learning-based travel time estimation approach designed for datasets with limited features. Our model uses only geocoordinates and departure times as inputs, and enhances predictive accuracy by incorporating derived nonlinear spatial features specifically, geodesic and Manhattan-style distances. We also investigate the effect of other derived input features, such as second-order polynomials of the basic inputs, to assess whether they contribute to performance improvement. This lightweight input representation enables rapid $O(n^{2})$ inference, making it suitable for VRP-scale scenarios. We also investigate the effects of architectural choices, including batch normalization, loss function variants, and input normalization, on model performance. When geodesic distance is added as an input feature, the validation loss (MSE) drops to about 14 % of the value obtained with coordinates alone, and the cumulative probability of achieving a relative error of at most 0.2 rises from 25 % to more than 50 %. These quantitative gains demonstrate high predictive accuracy and strong generalization to unseen OD pairs, suggesting that the proposed framework is a practical and effective solution for large-scale, real-world VRP applications.

The increasing frequency of urban flooding necessitates effective solutions for real-time navigation and predictive routing. This study presents IMPROVE Floodeye, an integrated mobile system designed to optimize vehicle navigation using internet of things IOT and ensemble algorithm. The system collects and analyzes real-time flood data from various sources, including weather reports, sensors, and user-generated inputs. By leveraging an ensemble algorithm that combines machine learning models, it predicts flood-prone areas and recommends alternative routes to ensure safe and efficient travel. The mobile application provides users with dynamic updates, visual flood maps, and adaptive route suggestions. Evaluation results demonstrate the system's accuracy in flood prediction and routing optimization compared to conventional navigation systems. The implementation of IMPROVE Floodeye offers a scalable and intelligent solution for urban flood management, enhancing commuter safety and reducing travel time. Based on the findings on the Floodeye system, the following recommendations are proposed to enhance its effectiveness, sustainability, and scalability. These include the integration of various types of sensors to more accurately measure flood levels and rainfall intensity, and the expansion of the routing scheme to cover a wider geographic area for improved data coverage. Additionally, incorporating community-based reporting can boost situational awareness and the reliability of flood monitoring. Collaborating with local government units (LGUs) is essential to support system deployment, integrate data with disaster response protocols, and foster public trust and adoption. Lastly, conducting long-term system evaluations is crucial for guiding future improvements and ensuring the continued sustainability of the project.

—In the modern world, frequent travel has become a necessity, with vehicles being the primary mode of transportation. Ensuring human safety while traveling is paramount. To address this, it is essential to adopt a combination of numerous static and dynamic parameters to achieve optimal route design in today’s complex transportation systems. This study introduces a methodology titled 'Multi-Factor Risk Assessment and Route Optimization for Safe Human Travel', which consists of three stages: Route Optimization, Risk Factor Analysis, and Data Collection. To assess the safety of various routes, a combination of dynamic and static factors is considered. These include traffic, weather, and road conditions, as well as vehicle-related factors such as type, age, and the surrounding road environment. By analyzing simulated data, the technique identifies potential risks and optimizes travel paths accordingly. For segmented routes, risk factors are calculated using both static and dynamic parameters, ensuring a comprehensive safety assessment. Prioritizing user safety, the system dynamically adjusts routes to offer the most cost-effective and safest travel options. This study lays a robust foundation for intelligent transportation systems, aimed at ensuring safer travel for users across a range of scenarios.

This study investigates the spatiotemporal characteristics and travel patterns of micro-electric vehicles (micro-EVs) by analyzing real-world trip data collected over three years from shared micro-EV services operating in three regions of South Korea. Individual trips were extracted from GPS-based trajectory data, and a network-based detour ratio was introduced to capture non-linear trip characteristics. In addition, a hierarchical clustering analysis was applied to identify heterogeneous micro-EV trip patterns. The results show that micro-EVs are predominantly used for short-distance urban trips, while a smaller but behaviorally distinct subset of trips demonstrates their capacity to support medium-distance travel under specific functional contexts. The clustering analysis identified six distinct trip pattern groups, ranging from dominant short-distance routine travel to less frequent patterns associated with adverse weather conditions and extreme detouring behavior. Overall, the findings suggest that micro-EVs function as a complementary urban mobility mode, primarily supporting localized travel while selectively accommodating extended-range and specialized trips. From a sustainability perspective, these findings highlight the role of micro-EVs as energy-efficient, low-emission alternatives to conventional passenger vehicles for short- and medium-distance urban trips. By empirically identifying heterogeneous and long-tailed micro-EV travel patterns, this study provides practical insights for sustainable urban mobility design and environmentally responsible transportation policies.

With the construction of urban roads, the increasing number of vehicles and the increasingly serious problem of traffic congestion. Accurate prediction of urban vehicle speed is of great significance to relieve traffic pressure and improve travel efficiency. This paper presents a city vehicle speed prediction model based on Long Short-Term Memory (LSTM), which can handle multi-source data, including historical traffic flow, weather conditions, holiday information, etc. As a deep learning model, LSTM experimental results show that it can capture long-term dependencies in traffic data in the case of multi-source data, and the proposed model has high accuracy and robustness in urban vehicle speed prediction.

Travel time reliability assessment has been widely used in recent years to evaluate the performance of transportation networks and measure the operation level of transportation systems. Weather, as one of the important factors influencing travel time reliability, affects the relationship between the supply and requirement of urban road networks. Considering the traffic characteristics under different traffic conditions, a study on the influence of weather on travel time reliability under different conditions is proposed to predict the probability of travelers completing their trips within the expected time under different weather conditions. Based on the urban road network data and cab trajectory data of Harbin city, this paper correlates the floating vehicle location with the road network information through a hidden Markov model to reduce the influence of vehicle trajectory errors on the calculation results of path travel time. To analyze the entire distribution of extreme travel time and its impact on the reliability of travel time under various traffic situations, it captures the tail features of the travel time distribution based on extreme value theory. Then, to increase the predictability of each quantile, it combines a deep-learning LSTM model and a quantile regression model to create a probabilistic travel time prediction model utilizing combined layers. The proposed model is compared with the linear quantile regression and neural network quantile regression models, and the model is evaluated in terms of point prediction results and probabilistic prediction results, respectively, to ensure the accuracy of predictions from the model. As a result, the prediction accuracy of the model in this paper is greatly improved, and the degree of violating quantile constraints is greatly reduced.

Winter weather can lead to dangerous driving conditions that cause reduced speeds and accidents, so rapid anti-icing and snow removal operations are necessary. This paper investigates the impact of winter road maintenance operations on traffic flow and individual vehicle maneuver. A comprehensive analysis framework is developed to incorporate connected vehicle data collected from snowplows and passenger vehicles, as well as road weather information, in assessing the impact of winter maintenance operations. By assigning vehicle trajectories to road segments and examining driving behavior at individual vehicle levels, we can better understand the impacts of snowplowing activities on traffic flow and vehicle maneuvering. The results indicate that winter road maintenance operations help restore reduced travel speeds due to snowy or icy conditions. Furthermore, while slow-moving snowplow trucks affect passenger vehicle longitudinal and lateral movements, statistical analysis reveals no significant association between snowplow activities and increased events of hard acceleration and traction control. The results of this study enable a better understanding of the potential impacts of winter maintenance operations not only on traffic flow, but also on vehicles adjacent to maintenance trucks.

Reliable short-term traffic forecasting is a core requirement for smart-city mobility management because congestion affects travel time reliability, signal control, routing, safety, and local air quality. This study develops a multi-location, multi-step forecasting framework for hourly urban traffic prediction in Trondheim, Norway, using data from six fixed traffic sensors combined with synchronized weather observations. The empirical design uses hourly passenger-car counts collected from December 2018 through January 2020, merged with meteorological variables and calendar-derived seasonality indicators. After data fusion, missing-value treatment, feature selection, and time-lag restructuring through a 24 h sliding window, a direct forecasting strategy is implemented for a 24-step horizon. The modeling framework compares a broad set of machine-learning and deep-learning regressors, beginning with 17 one-step-ahead candidates and then retaining the seven best-performing ensemble tree-based models for full multi-step forecasting, while recurrent neural networks are trained in parallel for comparison. The results show that interpretable ensemble tree-based methods dominate across the Trondheim case study. For the first forecasting step, Extra Trees achieves the highest accuracy in all six locations, with R2 values of 98.16%, 97.64%, 94.76%, 95.12%, 97.31%, and 96.22%, respectively. Across the full 24 h horizon, model accuracy declines gradually with increasing lead time, but ensemble tree-based models remain consistently stronger than the tested recurrent neural networks. Extra Trees and Random Forest perform especially well at longer horizons, whereas Histogram-Based Gradient Boosting Regressor and Light Gradient Boosting Machine emerge as the most reliable models overall across locations and forecast steps. The resulting framework is well aligned with the aims of urban development and smart-city research because it demonstrates how city-scale sensing infrastructure and interpretable predictive analytics can support proactive congestion management and operational transport planning under limited historical data.

Modeling the distribution of daily and hourly human mobility metrics is beneficial for studying underlying human travel patterns. In previous studies, some probability distribution functions were employed in order to establish a base for human mobility research. However, the selection of the most suitable distribution is still a challenging task. In this paper, we focus on modeling the distributions of travel distance, travel time, and travel speed. The daily and hourly trip data are fitted with several candidate distributions, and the best one is selected based on the Bayesian information criterion. A case study with online car-hailing data in Xi’an, China, is presented to demonstrate and evaluate the model fit. The results indicate that travel distance and travel time of daily and hourly human mobility tend to follow Gamma distribution, and travel speed can be approximated by Burr distribution. These results can contribute to a better understanding of online car-hailing travel patterns and establish a base for human mobility research.

No abstract available

Car sharing is one the pillars of a smart transportation infrastructure, as it is expected to reduce traffic congestion, parking demands and pollution in our cities. From the point of view of demand modelling, car sharing is a weak signal in the city landscape: only a small percentage of the population uses it, and thus it is difficult to study reliably with traditional techniques such as households travel diaries. In this work, we depart from these traditional approaches and we leverage web-based, digital records about vehicle availability in 10 European cities for one of the major active car sharing operators. We discuss which sociodemographic and urban activity indicators are associated with variations in car sharing demand, which forecasting approach (among the most popular in the related literature) is better suited to predict pickup and drop-off events, and how the spatio-temporal information about vehicle availability can be used to infer how different zones in a city are used by customers. We conclude the paper by presenting a direct application of the analysis of the dataset, aimed at identifying where to locate maintenance facilities within the car sharing operation area.

No abstract available

Car-following modeling is one of the most used approaches for road traffic modeling. It ensures a detailed overview of vehicles behavior at microscopic traffic modeling level, taking into account some primary parameters like velocity, acceleration/deceleration, the distance between vehicles etc. A big disadvantage of this model is that is single-lane oriented, studying the current vehicle behavior based only on vehicle ahead behavior. The purpose of this paper is to deliver a new car-following model capable to adapt to multiple lanes roads, where the followed vehicle can be changed at any time. In this case, a big challenge will be the integration of a new vehicle in the established car-following model. This study attempts to estimate these different cases of lane-change based on a Bayesian reasoning estimation, facilitating the new vehicle integration on the current lane. Results will show the advantage of having a multiple lanes road traffic overview in adopting a proper traffic strategy, from the possible routes that can be reached point of view, based on lane change drivers' decisions.

Car sharing is nowadays a popular means of transport in smart cities. In particular, the free-floating paradigm lets the customers look for available cars, book one, and then start and stop the rental at their will, within a specific area. This is done thanks to a smartphone app, which contacts a web-based backend to exchange information. In this paper we present UMAP, a platform to harvest the data freely made available on the web by these backends and to extract driving habits in cities. We design UMAP with two specific purposes. Firsty UMAP fetches data from car sharing platforms in real time. Secondly, it processes the data to extract advanced information about driving patterns and user's habits. To extract information, UMAP augments the data available from the car sharing platforms with mapping and direction information fetched from other web platforms. This information is stored in a data lake where historical series are built, and later analyzed using analytics modules easy to design and customize. We prove the flexibility of UMAP by presenting a case of study for the city of Turin. We collect car sharing usage data for over 50 days to characterize both the temporal and spatial properties of rentals, and to characterize customers' habits in using the service, which we contrast with public transportation alternatives. Results provide insights about the driving style and needs, which are useful for smart city planners, and prove the feasibility of our approach.

No abstract available

The estimation of traffic volume is an important part of urban transport management, and it is a necessary step for resource allocation and traffic flow optimization. In this study, we evaluate the performance of various different regression models of the relationship between traffic volume and a diverse collection of variables (e.g., temperature, precipitation (both snow and rain), weather, holiday status, and several temporal features such as year, month, day, hour, minute, and second). The paper utilizes a rich dataset of historical traffic volumes, combined with related environmental and temporal data. Missing values are handled and features standardized according to strict preprocessing methods. Linear Regression, Decision Tree Regression, Random Forest Regression, XGB-regression, & Support Vector Machine (SVM) regression are the regression models considered. Performance evaluation through metrics such as mean absolute error, mean squared error, and R-squared to validate the models for generalization and accuracy. Although the Random Forest Regressor and XGB-Regressor both receive fantastic R-squared values 97 and 90 respectively, suggesting mostly great predictive performance, the Decision Tree Regressor has an impressive R-squared score of 1.0, meaning it perfectly spaced the training data.

No abstract available

Traffic flow prediction is an important part of intelligent transportation systems (ITS). However, the performance of current traffic flow prediction methods does not meet the expectation. Weather factors such as precipitation in residential areas and tourist destinations affect traffic flow on the surrounding roads. In this paper, we attempt to take precipitation impact into consideration when predicting traffic flow. To realize this idea, we propose a deep traffic flow prediction architecture by introducing a deep bi-directional long short-term memory model, precipitation information, residual connection, regression layer and dropout training method. The proposed model has good ability to capture the deep features of traffic flow. Besides, it can take full advantage of time-aware traffic flow data and additional precipitation data. We evaluate the prediction architecture on the dataset from Caltrans Performance Measurement System (PeMS) with precipitation data from California Data Exchange Center (CDEC) and the dataset from KDD Cup 2017. The experiment results demonstrate that the proposed model for traffic flow prediction obtains high accuracy and generalizes well compared with other models.

This paper examines the impact of meteorological conditions on traffic flow and its speed in Prague, focusing on variations in precipitation, and temperature to evaluate emission data. We utilize data from strategic traffic detectors operated by the municipality of Prague and sensors from the Czech Hydrometeorological Institute, ensuring the use of high-quality, verified datasets. The study explores the relationships between weather fluctuations, traffic dynamics, and emission parameters. For this purpose, we are using knowledge graphs for relationship detection and fuzzy inference systems for the prediction of changes in emissions depending on various conditions (traffic situation and weather). The findings indicate that weather conditions and traffic flow significantly influence the emission behavior depending on the geographical location and parameters of the investigated urban area. This research enhances our understanding of human and goods mobility within urban settings, specifically in the context of city logistics.

No abstract available

No abstract available

No abstract available

Traffic flow forecasting is essential for managing congestion, improving safety, and optimizing various transportation systems. However, it remains a prevailing challenge due to the stochastic nature of urban traffic and environmental factors. Better predictions require models capable of accommodating the traffic variability influenced by multiple dynamic and complex interdependent factors. In this work, we propose a Graph Neural Network (GNN) framework to address the stochasticity by leveraging adaptive adjacency matrices using log-normal distributions and Coefficient of Variation (CV) values to reflect real-world travel time variability. Additionally, weather factors such as temperature, wind speed, and precipitation adjust edge weights and enable GNN to capture evolving spatio-temporal dependencies across traffic stations. This enhancement over the static adjacency matrix allows the model to adapt effectively to traffic stochasticity and changing environmental conditions. Furthermore, we utilize the Adaptive Conformal Prediction (ACP) framework to provide reliable uncertainty quantification, achieving target coverage while maintaining acceptable prediction intervals. Experimental results demonstrate that the proposed model, in comparison with baseline methods, showed better prediction accuracy and uncertainty bounds. We, then, validate this method by constructing traffic scenarios in SUMO and applying Monte-Carlo simulation to derive a travel time distribution for a Vehicle Under Test (VUT) to reflect real-world variability. The simulated mean travel time of the VUT falls within the intervals defined by INRIX historical data, verifying the model’s robustness.

Traffic information is crucial for managing transportation and city planning, but obtaining national-scale data is difficult due to privacy concerns. Consequently, most current traffic datasets have limitations in terms of time and location coverage, leading to a lack of comprehensive public access to national traffic data. To address this issue, a multi-source highway traffic dataset has been created, featuring 2042 sensors in New Zealand over a 9-year period with 15-minute intervals and accompanying metadata. The dataset includes data of both light-duty and heavy-duty vehicles, as well as weather information like temperature and precipitation. This dataset has diverse potential research applications such as traffic flow prediction and congestion management.

No abstract available

Traffic congestion has become more concerning as the number of cars has rapidly increased in urban environments, resulting in dangerous pollution levels amid global warming. The effect of weather factors on traffic flow is evident thus studying this relationship can improve traffic prediction, which can help us to prevent or dissolve traffic congestion in smart cities where intelligent transportation systems have been deployed. In this paper, a new method, the Weather Based Traffic Analysis (hereafter WBTA) has been proposed to investigate the temporal correlations between traffic flow and exogenous weather factors, at different frequencies and time intervals. This can help in the proper selection of the most important weather factors which influence traffic flow behavior, in addition to the duration of that effect. This can help in choosing input parameters of a multi-input traffic prediction model. Several weather factors are analyzed, however, for the sake of this paper, the precipitation factor is discussed.

Abstract Severe weather events pose a significant threat to transportation networks. This research analyzes and discusses the impact of precipitation, temperature, visibility and wind speed on hourly weekday traffic flow volume in Atlanta, Georgia. The study involves the following: determine weather variables that affect traffic volume, develop a machine learning technique to derive decision rules based on weather and traffic volume, and create a web-based decision support visualization tool using the analyzed results. The relationship between extreme weather events and traffic volume was investigated by comparing traffic volume between a base case scenario and an extreme weather scenario. Data from 48 Automatic Traffic Recorder (ATR) sites around Atlanta, GA, USA and hourly precipitation data from 4 climate measurement stations were used to conduct this study. The spatiotemporal relationships between traffic volume and weather variables were analyzed individually and evaluated using a non-parametric statistical test. A machine learning technique is applied to derive decision rules that result in reduction in traffic volume. Results show significant impacts on traffic volume from visibility, precipitation and temperature and helps in isolating hours in a typical weekday when such impacts are felt. A decision support tool was also developed to visualize traffic volume and weather interactions. The data-driven insights from this analysis is applicable to transportation planners, centralized traffic control rooms and urban infrastructure decision makers.

No abstract available

This study investigates the prediction and mitigation of the phenomenon of traffic flow breakdown when affected by varying weather conditions. First, the probability of breakdown occurrence is examined using a survival analysis approach to obtain distributions of pre-breakdown flow rates under different weather conditions. Second, pre-breakdown flow rate distributions were applied in breakdown prediction for the implementation of breakdown mitigation strategies. In the first part, a set of data from the network of Kansas City was used to demonstrate the applicability of the Kaplan–Meier Product Limit method to estimating the breakdown probability under various weather conditions. Then, using simulated data on the network of Chicago, the K-M approach was used again to obtain survival likelihood distributions, which in turn yield breakdown probability, for 13 different weather cases as combinations of weather categories for different levels of visibility, rain, and snow precipitation. In the second part, continuing with the simulated data, dynamic speed limits (DSL) were applied to demonstrate the effectiveness of the prediction method presented. A sensitivity analysis of the threshold probability and upstream distance at which DSL should be implemented was performed for clear and inclement weather conditions. In clear weather the performance of the strategy is better at a lower probability threshold and farther upstream location, whereas in inclement weather the performance is better at a lower probability threshold and closer upstream location. The paper demonstrates the effect of changing weather conditions on the likelihood of breakdown occurrence and the implementation of breakdown mitigation strategies.

No abstract available

No abstract available

No abstract available

No abstract available

This study proposes a methodical approach to model desired speed distributions under different road-weather and traffic conditions followed by identification of road-weather conditions with potentially higher safety risks in rural divided highways located in extremely cold regions. Desired speed distributions encompassing unique combinations of adverse road-weather and traffic conditions are modelled as normal distributions characterized by their means and standard deviations formulated based on two principal statistical theorems and techniques i.e., Central Limit Theorem and Minimum Variance Unbiased Estimation. Combination of the precipitation conditions, road surface conditions, time of the day, temperature, traffic flow and the heavy vehicle percentage at the time of travel were considered in defining the combinations of road-weather and traffic conditions. The findings reveal that simultaneous occurrence of particular precipitation and pavement conditions significantly affect the characteristics of the desired speed distribution and potentially expose drivers to elevated safety risks. Jurisdictions experiencing extreme road-weather conditions may adapt the proposed methodology to assess speed behaviour under different road-weather conditions to establishing and deploying weather-responsive traffic management strategies such as variable speed limit to regulate speeding and improve traffic safety in winter.

No abstract available

In the context of rapidly growing city road networks, understanding complex traffic patterns and implementing effective safety monitoring through advanced Transportation Cyber-Physical Systems (T-CPS) has become increasingly challenging. This involves understanding spatial relationships and non-linear temporal associations. Accurately predicting traffic in such scenarios, particularly for long-term sequences, is challenging due to the complexity of the data. Traditional ways of predicting traffic flow use a single fixed graph structure based on location. This structure does not consider possible correlations and cannot fully capture long-term temporal relationships among traffic flow data, thereby limiting the system ability to ensure safety and reliability. To address this challenge, we propose a novel traffic prediction framework called Attention-based Spatio-temporal Multi-scale Graph Convolutional Recurrent Network (ASTMGCNet). This study introduces a novel framework designed to improve prediction accuracy in dynamic urban traffic systems by effectively capturing complex spatio-temporal correlations through multi-scale feature extraction and attention mechanisms. ASTMGCNet records changing features of space and time by combining Gated Recurrent Units (GRU) and Graph Convolutional Networks (GCN). Its design incorporates multi-scale feature extraction and dual attention mechanisms, effectively capturing informative patterns at different levels of detail. This strategic design allows ASTMGCNet to effectively capture complex spatio-temporal correlations within traffic sequences, enhancing prediction accuracy. We have tested this method on two different real-world datasets and found that ASTMGCNet predicts significantly better than other methods, demonstrating its potential to advance traffic flow prediction and improve safety and reliability in T-CPS applications.

Accurate traffic flow prediction is essential for applications like transport logistics but remains challenging due to complex spatio-temporal correlations and non-linear traffic patterns. Existing methods often model spatial and temporal dependencies separately, failing to effectively fuse them. To overcome this limitation, the Dynamic Spatial-Temporal Trend Transformer DST2former is proposed to capture spatio-temporal correlations through adaptive embedding and to fuse dynamic and static information for learning multi-view dynamic features of traffic networks. The approach employs the Dynamic Trend Representation Transformer (DTRformer) to generate dynamic trends using encoders for both temporal and spatial dimensions, fused via Cross Spatial-Temporal Attention. Predefined graphs are compressed into a representation graph to extract static attributes and reduce redundancy. Experiments on four real-world traffic datasets demonstrate that our framework achieves state-of-the-art performance.

Accurately predicting road traffic flows is a primary challenge in the development of smart cities, providing a scientific basis and reference for urban planning, construction, and traffic management. Road traffic flow is influenced by various complex features, including temporal and weather conditions, which introduce challenges to traffic flow prediction. To enhance the accuracy of traffic flow prediction and improve the adaptability across different weather conditions, this study introduced a traffic flow prediction model with explicit consideration of weather factors including temperature, rainfall, air quality index, and wind speed. The proposed model utilized grey relational analysis (GRA) to transform weather data into weighted traffic flow data, expanded input variables into a new data matrix, and employed one-dimensional convolutional neural networks (CNNs) to extract valuable feature information from these input variables, as well as bidirectional long short-term memory (BiLSTM) to capture temporal dependencies within the time-series data. Bayesian optimization was employed to fine-tune the hyperparameters of the model, offering advantages such as fewer iterations, high efficiency, and fast speed. The performance of the proposed prediction model was validated using the traffic flow data collected at an intersection in China and on the M25 motorway in the United Kingdom. The results demonstrated the effectiveness of the proposed model, achieving improvements of at least 9.0% in MAE, 2.8% in RMSE, 2.3% in MAPE, and 0.06% in R2 compared to five baseline models.

Forecasting the traffic flows is a critical issue for researchers and practitioners in the field of transportation. However, it is very challenging since the traffic flows usually show high nonlinearities and complex patterns. Most existing traffic flow prediction methods, lacking abilities of modeling the dynamic spatial-temporal correlations of traffic data, thus cannot yield satisfactory prediction results. In this paper, we propose a novel attention based spatial-temporal graph convolutional network (ASTGCN) model to solve traffic flow forecasting problem. ASTGCN mainly consists of three independent components to respectively model three temporal properties of traffic flows, i.e., recent, daily-periodic and weekly-periodic dependencies. More specifically, each component contains two major parts: 1) the spatial-temporal attention mechanism to effectively capture the dynamic spatialtemporal correlations in traffic data; 2) the spatial-temporal convolution which simultaneously employs graph convolutions to capture the spatial patterns and common standard convolutions to describe the temporal features. The output of the three components are weighted fused to generate the final prediction results. Experiments on two real-world datasets from the Caltrans Performance Measurement System (PeMS) demonstrate that the proposed ASTGCN model outperforms the state-of-the-art baselines.

Traffic flow forecasting is a classical spatio-temporal data mining problem with many real-world applications. Recently, various methods based on Graph Neural Networks (GNN) have been proposed for the problem and achieved impressive prediction performance. However, we argue that the majority of existing methods disregarding the importance of certain nodes (referred to as pivotal nodes) that naturally exhibit extensive connections with multiple other nodes. Predicting on pivotal nodes poses a challenge due to their complex spatio-temporal dependencies compared to other nodes. In this paper, we propose a novel GNN-based method called Spatio-Temporal Pivotal Graph Neural Networks (STPGNN) to address the above limitation. We introduce a pivotal node identification module for identifying pivotal nodes. We propose a novel pivotal graph convolution module, enabling precise capture of spatio-temporal dependencies centered around pivotal nodes. Moreover, we propose a parallel framework capable of extracting spatio-temporal traffic features on both pivotal and non-pivotal nodes. Experiments on seven real-world traffic datasets verify our proposed method's effectiveness and efficiency compared to state-of-the-art baselines.

Abstract: Given the complexity of urban transportation networks and the multiple variables that might affect journey times, estimating the length of cab rides in New York City (NYC) is a difficult process. In this study, we provide a unique method for resolving this issue that makes use of machine learning techniques and a wide range of attributes gleaned from taxi trip data. We start by gathering a sizable collection of historical records of NYC taxi trips,providing specifics like pick-up and drop-off points, timestamps, and lengths of trips. To deal with outliers, missing values, and geographical and temporal irregularities, we preprocess the data. Furthermore, we design a broad range of characteristics, such as geographic coordinates, time of day, day of the week, and weather conditions, to capture the spatial, temporal, and contextual elements of each journey. Then, using gradient boosting methods, we create a prediction model that efficiently uncovers the intricate patterns seen in the data. We carefully adjust the model's hyperparameter to enhance performance and use cross-validation techniques to guarantee resilience. In addition, we apply ensemble techniques to enhance prediction precision and minimise model bias. We conduct lengthy tests on a held-out test set and compare the performance of our model to a number of baseline techniques frequently employed in triptime prediction in order to assess the efficacy of our suggested strategy. The outcomes show that our strategy works better than the competition, with lower prediction errors and higher accuracy. We also do interpretability assessments to learn more about the variables that have the most impact on estimates of trip time. Our results demonstrate the potential of feature engineering and M L approaches for precise and trustworthy taxi trip length prediction in NYC. The suggested method not only helps taxi service companies by allowing them to more accurately predict journey lengths, but it also improves customer experience by giving more precise travel time estimates. Additionally, our approach may be used as a starting point for future studies in the field of urban transportation prediction, enabling better efficiency and planning in urban mobility networks.

Travel time prediction is crucial in developing mobility on demand systems and traveller information systems. Precise estimation of travel time supports the decision-making process for riders and drivers who use such systems. In this paper, static travel time for taxi trip trajectories is predicted by applying isolated XGBoost regression models to a set of identified inlier and extreme-conditioned trips and the results are compared with other existing best models in this context. XGBoost uses an ensemble of decision trees and is robust to outliers and thus it is believed to perform well on time series predictions. We show that, compared to other existing best models, XGB-IN (XGBoost prediction model of in-lier trips) model prediction values reduce mean absolute error as well as root mean squared error and exhibit impressive correlation with actual travel time values while XGB-Extreme model is able to provide reasonably accurate prediction results for a set of extreme-conditioned trips with shorter actual time durations. We demonstrate the achievability of travel time prediction with XGBoost regression and show that our approach is applicable to large-scale data and performs well in predicting static travel time.

No abstract available

One of the crucial problems for taxi drivers is to efficiently locate passengers in order to increase profits. The rapid advancement and ubiquitous penetration of Internet of Things (IoT) technology into transportation industries enables us to provide taxi drivers with locations that have more potential passengers (more profitable areas) by analyzing and querying taxi trip data. In this paper, we propose a query processing system, called Distributed Profitable-Area Query (DISPAQ) which efficiently identifies profitable areas by exploiting the Apache Software Foundation’s Spark framework and a MongoDB database. DISPAQ first maintains a profitable-area query index (PQ-index) by extracting area summaries and route summaries from raw taxi trip data. It then identifies candidate profitable areas by searching the PQ-index during query processing. Then, it exploits a Z-Skyline algorithm, which is an extension of skyline processing with a Z-order space filling curve, to quickly refine the candidate profitable areas. To improve the performance of distributed query processing, we also propose local Z-Skyline optimization, which reduces the number of dominant tests by distributing killer profitable areas to each cluster node. Through extensive evaluation with real datasets, we demonstrate that our DISPAQ system provides a scalable and efficient solution for processing profitable-area queries from huge amounts of big taxi trip data.

No abstract available

No abstract available

No abstract available

The primary objective of this study was to investigate how trip pattern variables extracted from large-scale taxi GPS data contribute to the spatially aggregated crashes in urban areas. The following five types of data were collected: crash data, large-scale taxi GPS data, road network attributes, land use features and social-demographic data. A data-driven modeling approach based on Latent Dirichlet Allocation (LDA) was proposed for discovering hidden trip patterns from a taxi GPS dataset, and a total of fifty trip patterns were identified. The collected data and the identified trip patterns were further aggregated into167 ZIP Code Tabulation Areas (ZCTA). Random forest technique was used to identify the factors that contributed to total, PDO and fatal-plus-injury crashes in the selected ZCTAs during the study period. Geographically weighted Poisson regression (GWPR) models were then developed to establish a relationship between the crashes and the contributing factors selected by the random forest technique. Comparative analyses were conducted to compare the performance of the GWPR models that considered traditional traffic exposure variables only, trip pattern variables only, and both traditional exposure and trip pattern variables. The model specification results suggest that the trip pattern variables significantly affected the crash counts in the selected ZCTAs, and the models that considered both the traditional traffic exposure and the trip pattern variables had the best goodness-of-fit in terms of the lowest MAD and AICc values.

Crowdsourcing applications are proven to be a promising tool to gather valuable information, which can be used for a wide range of tasks, such as ensuring public safety. Traffic data collected using these applications have been used for efficient evacuation planning in large cities. In this paper, we propose to use regression-based machine learning methods to predict hourly taxi rides for a given location in a target day of week and month. The presented method can be used for the following purposes: 1) Predicting the number of taxi rides for a given location at a given time, 2) Identifying hot spots in a city, 3) Getting a rough count of the population density at a given location at a targeted hour, and 4) Planing evacuation routes for possible disasters. The presented approach has potential use for resource planning and evacuation in large cities. The Taxi and Limousine Commission (TLC) trip record data collected from 2017 to 2018 was used for this experiment. It was found that random forest regression can successfully predict hourly taxi drop-offs for a given taxi zone as well as for the entire city of New York.

Both taxi and Internet-based ride-hailing services are important public transportation options. While these services provide a similar functionality, their quantitative patterns are difficult to describe and analyze without a big data approach. In this study, we present a data-driven analysis for taxi and ride-hailing services with a case study in Chengdu, China. Both a taxi GPS dataset and an Internet-based ride-hailing trip dataset provided by Didi Chuxing, the largest transportation network company in China, are used in this study. Our analysis is based on four aspects, i.e., temporal patterns, spatial patterns, spatio-temporal patterns, and traveling distance patterns. It is found that both taxi and ride-hailing services exhibit the densification power law with a space-time graph model. The ride-hailing service has a preference for more concentrated pickup and dropoff hotspots and trips with a longer distance. The observed results indicate that the ride-hailing service partially plays the role of taxis, but it also has its own characteristics as a new transport option. The findings in this study would be helpful for the government to better regulate the operation of both taxi and ride-hailing services.

No abstract available

Statistical distribution of taxi demand is essential for modeling the dynamics of taxi services. An accurate demand prediction not only helps the drivers lessen their searching time but also helps the passengers shorten their waiting time. Moreover, the temporal distribution of taxi demand is a critical component in traffic simulation. Obviously, a simulator needs to know how many new taxi pickup events to schedule after the others. In most studies, a poisson distribution is often used to model the temporal distribution of taxi pickups. However, this assumption has mostly been used without validation with empirical data. Therefore, it is unclear whether such an assumption is appropriate for modeling the statistical distribution of taxi demand. In this study, we characterize the temporal distribution of taxi pickups based on real taxi trip data from Bangkok, Thailand, and Chicago, IL, USA. It is shown that, in most cases, the poisson distribution is not suitable for modeling the temporal distribution of taxi pickups. On the contrary, this study demonstrates that a geometric distribution is more appropriate in modeling the temporal distribution of taxi pickups. To our knowledge, this has not been discovered in any prior studies.

Abstract Public buses and taxis play crucial roles in urban transportation. Ensuring their safety is of paramount importance to develop sustainable communities. This study investigated the significant factors contributing to the injury severity of bus–taxi crashes, using the crash data recorded by the police in Hong Kong from 2009 to 2019. To account for the unobserved heterogeneity, the random parameters logistic model with heterogeneity in means was elaborately developed. The results revealed that taxi driver age, bus age, traffic congestion, and taxi driver behavior had significantly heterogeneous effects on the injury severity of bus–taxi crashes and that the mean value of the random parameter for severe traffic congestion was likely to increase if the taxi’s age was <5 years. Taxi driver gender, rainfall, time of day, crash location, bus driver behavior, and collision type were found to significantly affect the bus–taxi crash severity. Specifically, female taxi drivers, old taxis, rainfall, midnight, improper manipulation of bus and taxi drivers, head-on and sideswipe collision types, and non-intersections were associated with a higher likelihood of fatal and severe crashes. Based on our findings, targeted countermeasures were proposed to mitigate the injury severity of bus–taxi crashes.

ABSTRACT A comprehensive understanding of human mobility patterns in urban areas is essential for urban development and transportation planning. In this study, we create entropy-based measurements to capture the geographical distribution diversity of trip origins and destinations. Specifically, we develop origin-entropy and destination-entropy based on taxi and ride-sharing trip records. The origin-entropy for a given zone accounts for all the trips that originate from this zone and calculates the level of geographical distribution diversity of these trips’ destinations. Likewise, the destination-entropy for a given zone considers all the trips that end in this zone and calculates the level of geographical distribution diversity of these trips’ origins. Furthermore, we have created an interactive geovisualization that enables researchers to delve into and juxtapose the spatial and temporal dynamics of origin and destination entropy, in conjunction with trip counts for both origins and destinations. Results indicate that entropy-based measurements effectively capture shifts in the diversity of trips’ geographical origins and destinations, reflecting changes in travel decisions due to major events like the COVID-19 pandemic. These measurements, alongside trip counts, offer a more comprehensive understanding of urban human flows.

No abstract available

ABSTRACT This paper presents a new path-finding problem to ensure reliable energy consumption for electric vehicles (EVs) under rainfall conditions. The objective function of the proposed model aims to find a reliable path that minimises energy consumption while ensuring a certain probability of completing the trip without exhausting a given battery energy budget. By considering the influence of adverse weather conditions at different periods, the existing model is expanded. To address the non-additivity and non-linearity characteristics of the optimisation model, an enhanced heuristic algorithm is proposed, incorporating inequality techniques, the K-shortest algorithm, and path-updating strategies. Lastly, the proposed algorithm is validated using Hong Kong’s grid-based road network as a case study, which demonstrates the correctness and effectiveness of the algorithm. The results indicate that by considering adverse weather conditions, the estimation of energy consumption can be significantly improved in terms of accuracy, achieving more efficient and reliable optimal path recommendations.

No abstract available

The frequent occurrence of extreme weather events (EWEs) in recent years has posed major hazards to urban transportation as well as socioeconomic impacts. A quantitative evaluation of the urban transportation resilience to minimize the impact caused by EWEs becomes critical to the rapid recovery of urban transportation after disasters. However, there is, generally, a lack of reliable data sources to monitor urban transportation performance under EWEs. This empirical study proposes a performance indicator (displacement) and quantitative method for evaluating the urban transportation performance under EWEs based on bus GPS trajectory datasets. Furthermore, the transportation resilience of it is quantified, and the variation is compared across temporal and spatial dimensions. The method is applied in a case study of Fuzhou, China, under rainstorm events. The results show that the Gulou and Jinan subareas have the highest transportation resilience during the yellow and red rainstorm warnings. By formulating an emergency plan and taking mitigation measures, the transportation performance in the Jinan subarea during the red rainstorm warning was improved by 36% compared to the yellow rainstorm warning. The empirical study not only fills the knowledge gap for quantifying the transportation resilience across the geographical boundary under rainstorm events, but also estimates the operation status of the road network. The results will help policymakers prioritize the resource distribution and develop effective policies or measures to further improve transportation resilience in the city.

This study focuses on the main urban area of Yangzhou City and conducts a quantitative comparative analysis of traffic accessibility during normal weather and extreme precipitation conditions (typhoon) based on GPS trajectories of buses. From both temporal and spatial dimensions, it comprehensively examines the impact of extreme precipitation on bus travel speed, travel time, and the commuting range of residents in the main urban area of Yangzhou City. (1) Through the mining and analysis of multi-source heterogeneous big data (bus GPS trajectory data, bus network data, rainfall remote sensing data, and road network data), it is found that the rainstorm weather greatly affects the average speed and travel time of buses. In addition, when the intensity of heavy rainfall increases (decreases), the average bus speed and travel time exhibit varying degrees of spatio-temporal change. During the morning and evening rush hour commuting period of rainstorm weather, there are obvious differences in the accessibility change in each typical traffic community in the main urban area of Yangzhou city. In total, 90% of the overall accessibility change value is concentrated around −5 min~5 min, and the change range is concentrated around −25~10%. (2) To extract the four primary traffic districts (Lotus Pond, Slender West Lake, Jinghua City, and Wanda Plaza), we collected Points of Interest (POI) data from Amap and Baidu heat map, and a combination analysis of the employment–residence ratio model and proximity methods was employed. The result show that the rainstorm weather superimposed on the morning peak hour has different degrees of impact on the average speed of the above-mentioned traffic zones, with the most obvious impact on the Lotus Pond and the smallest impact on Wanda Plaza. Under the rainstorm weather, the traffic commute in the main urban area of Yangzhou in the morning and evening peak hour is basically normal. The results of this paper can help to quantify the impact of typhoon-rainstorm weather events on traffic commuting in order to provide a scientific basis for the traffic management department to effectively prevent traffic jams, ensure the reliability of the road network, and allow the traffic management department to more effectively manage urban traffic.

ABSTRACT Traffic congestion estimation on urban road segments is crucial to traffic management. Considering the heterogeneous impact of dynamic critical bottleneck (e.g., arterials) on congestion diffusion, this study proposes a traffic congestion estimation method by using GPS trajectory data. At first, the process of congestion diffusion is modeled by percolation theory, and the critical threshold ${q_c}\left(T \right)$qcT in time interval $T$T is inferred to represent the network-wide traffic states. Then, ${q_c}\left(T \right)$qcT is utilized as the baseline to characterize the heterogeneous impact of dynamic critical bottlenecks on congestion diffusion. Finally, the Systemic Congestion Index (SCI) is generated to estimate segment-based congestion intensity. Investigations revealed that compared with the speed, relative velocity (RV), Travel Time Index (TTI), and the ground truth data (i.e. occupancy), the proposed method can capture the spatial-temporal variation of congestion. Moreover, the reliability of SCI is verified by the Dynamic Time Warping algorithm (DTW) and a sensitivity analysis.

In recent years, the use of motorcycles has witnessed a remarkable surge in urban areas, paralleled by a growing demand for research and development in the motorcycle industry. Furthermore, the widespread adoption of GPS-enabled devices over the last few decades has opened up exciting possibilities, particularly in the realm of data analysis, where motorcycle GPS data has emerged as a valuable resource for various applications. This article presents a novel methodology for estimating the travel duration of powered two-wheelers (PTWs) in heterogeneous traffic using GPS data generated by motorcycles on urban road networks. The proposed methodology has the potential to offer valuable insights into the behavior of PTWs in heterogeneous traffic environments. By analyzing Big Data generated by GPS-based trajectory data, researchers can identify areas with high motorcycle density and pinpoint potential bottlenecks that impact travel times. Temporal data storing with bearing information in hexagonal shards called “bubbles” enables researchers to utilize Big Data more efficiently. Spatial transformation, Kalman filtering, and map-matching of the trajectory data significantly enhance the quality of the data. In this study, the 10-minute interval is performed as optimal for estimating travel time with 4.3% MAPE. Furthermore, combining historical bubble data with a 0.35 scale factor improves MAPE by 9.6%. Despite the limitations, not only is the transferability of this methodology noteworthy, but it is also opening the door to broader applications in diverse urban settings.

Surrogate safety measures (SSMs) have drawn considerable attention in the field of traffic safety. Previous studies have characterized SSMs based on road geometries and conflict types, exploring various SSM thresholds. Adverse weather conditions have been examined for their impact on safety events estimated from surrogate metrics, however, limited research focuses on discrepancies in SSM thresholds under various weather characteristics. This study investigated specific thresholds for modified time to collision (MTTC), deceleration rate to avoid a crash (DRAC), and single-step probabilistic driving risk field (S-PDRF) under clear and light rainy weather conditions. A total of 1,048 rear-end events in the expressway (724 clear and 324 rainy) were obtained from the CitySim dataset. A statistical test was conducted to examine the significance of three indicators under two weather scenarios. Subsequently, the peak over threshold (POT) and bimodal histogram threshold methods were employed to recommend appropriate thresholds. The results indicate that thresholds of MTTC and DRAC metrics have statistically significant differences under two weather scenarios. Specifically, the recommended MTTC thresholds were 4 s for rainy conditions and 2.3 s for clear conditions. Similarly, the DRAC thresholds were 3.2 m/s2 for clear weather and 2.4 m/s2 for rainy weather. Conversely, the threshold of S-PDRF was statistically insignificant between two weather conditions, suggesting a single threshold is appropriate for this measure regardless of weather. These findings can inform the design of active safety systems and traffic safety policies by suggesting different safety thresholds for triggering safety systems under clear and rainy conditions, thereby reducing false alarms.

Massive Global Positioning System (GPS) trajectory datasets are being produced owing to the advances in mobile sensors, the Internet, and GPS devices. Accurately inferring travel modes from GPS trajectory data can be helpful in transportation planning and modeling, infrastructure design, etc. However, adverse factors such as data noise, differences in sampling rate, and inadequate features have a negative impact on the results of travel mode classification. In this paper, to address such issues, we first propose a preprocessing workflow, which includes data cleaning, segmentation, and resampling, to preprocess raw trajectories. Then, we add new features related to the road and bus stop information for travel mode classification using an XGBoost (eXtreme Gradient Boosting) classifier, along with various basic features of the trajectories. We conducted a set of experiments on the GeoLife dataset using a group of state-of-the-art methods. The results showed that the proposed methods can improve the classification accuracy by using all the classifiers we compared and the classification accuracy using the XGBoost classifier can reach a maximum of 90.41%.

Understanding the sideslip risks of various trajectory patterns, as well as the impact of rainfall on them, is critical for improving road safety. However, the lack of precise classification indicators hampers systematic analysis of the variations in vehicle trajectory patterns. To address this, this study proposes a parameterized classification method for trajectories on curved segments, employing the radius and offset of the trajectory as the primary classification features and dividing the trajectories into nine patterns. These patterns represent variations from smaller to larger radii and inside to outside lane offsets, reflecting different driving behaviors and vehicle stability during vehicle cornering. Concurrently, the friction coefficient utilization rate is used to effectively compare vehicles’ sideslip risk under different weather conditions. Based on this, we construct a framework using computer vision technology for automatically identifying trajectory patterns and measuring sideslip risk. We conducted an empirical study on a highway-curved segment with high sideslip risk in China and collected two datasets under clear and rainy conditions for analysis. The classification results show that the proposed method can effectively classify trajectories according to nine trajectory patterns. Comparative analysis reveals that vehicle trajectories in both the inside and outside lanes are notably more affected by rainfall compared to the middle lane. Meanwhile, trucks demonstrate a higher susceptibility to rainfall than cars. In addition, the analysis of the sideslip risk for different trajectory patterns discovers several high-risk patterns. This study provides an effective approach for monitoring and analyzing the sideslip risk on curved segments, thereby contributing to the enhancement of road design and traffic safety management.

Abstract Pre-planned events such as constructions or special events lead to road capacity reductions and create bottlenecks in the traffic network. The traffic impact of such events goes beyond local areas, as informed drivers may detour to alternative corridors and consequently the traffic congestion may divert or propagate to other corridors. Due to the lack of real observation data, traditional traffic impact analyses are typically based on simulation models, fixed-location sensor data or survey questionnaires. In this research, we use high-resolution vehicle trajectory data collected via a smartphone app, which is capable of keeping track of individual driver’s behavior before and after road capacity reduction, to investigate travelers’ behavioral responses to pre-planned events and the contribution factors. For this purpose, a functional data analysis (FDA) approach-based clustering method is firstly proposed to cluster trajectory data and identify detour patterns, and two logistic and a least absolute shrinkage and selection operator (LASSO) regression models are used to explain drivers’ detour behavior choice for each pattern with spatial and temporal features of interest. A case study based on a lane closure event on MoPac expressway in Austin, TX is used as an example in this research. The case study demonstrates that: (1) the freeway capacity reduction triggered heterologous behavior responses, (2) driver detour behavior exhibits three major patterns and (3) each detour pattern highly depends on spatial features such as trip length, distance to freeway entrance and distance to other alternative freeways, in addition to the temporal features when the trip happens.

Globally, smart cities, infrastructure, and transportation have led to a rise in vehicle numbers, resulting in an increasing number of problems. This includes problems such as air pollution, noise pollution, high energy consumption, and people’s health. A viable solution to these problems is carpooling, which involves sharing vehicles between people going to the same location. As carpooling solutions become more popular, they need to be implemented efficiently. Data analytics can help people make informed decisions when selecting a ride (Car or Bus). We applied machine learning algorithms to select the desired ride (Car or Bus) and used feature ranking algorithms to identify the foremost traits for selecting the desired ride. Based on the performance evaluation metric, 11 classifiers were used for the experiment. In terms of selecting the desired ride, Random Forest performs best. Using ten-fold cross-validation, we obtained a sensitivity of 87.4%, a specificity of 73.7%, an accuracy of 81.0%, a sensitivity of 90.8%, a specificity of 77.6%, and an accuracy of 84.7% using leave-one-out cross-validation. To identify the most favorable characteristics of the Ride (Car or Bus), the recursive elimination of features algorithm was applied. By identifying the factors contributing to users’ experience, the service providers will be able to rectify those factors to increase business. It has been determined that the weather can make or break the user experience. This model will be used to quantify and map intrinsic and extrinsic sentiments of the people and their interactions with locality, socio-economic conditions, climate, and environment.

No abstract available

Abstract: This research presents a predictive logistics framework designed to enhance disaster response efficiency during earthquakes and floods by leveraging integrated traffic, weather, and supply chain data. The proposed system combines deep learning models—such as LSTMs and CNNs—for disaster impact forecasting, with dynamic graph models for traffic disruption prediction and probabilistic approaches for supply chain risk assessment. A multi-objective optimization engine, powered by genetic algorithms and reinforcement learning, dynamically allocates resources, plans routes, and prioritizes delivery based on evolving disaster scenarios. Real-time data from GPS networks, meteorological sources, and inventory systems are processed and visualized via an interactive GIS-enabled dashboard. The system demonstrates significant improvements in delivery time, coverage ratio, and operational resilience compared to static models. Designed for scalability and modularity, the framework offers a powerful decision-support tool for governments, NGOs, and emergency responders aiming to minimize humanitarian impact and maximize logistical effectiveness during compound natural disasters. Keywords: disaster logistics, predictive modeling, earthquake response, flood response, traffic disruption, weather forecasting, supply chain resilience, LSTM, reinforcement learning, multi-objective optimization, humanitarian logistics, GIS, emergency response planning, dynamic routing, real-time decision support

Recovering intermediate missing GPS points in a sparse trajectory, while adhering to the constraints of the road network, could offer deep insights into users’ moving behaviors in intelligent transportation systems. Although recent studies have demonstrated the advantages of achieving map-constrained trajectory recovery via an end-to-end manner, they still face two significant challenges. First, existing methods are mostly sequence-based models. It is extremely hard for them to comprehensively capture the micro-semantics of individual trajectory, including the information of each GPS point and the movement between two GPS points. Second, existing approaches ignore the impact of the macro-semantics, i.e., the road conditions and the people's shared travel preferences reflected by a group of trajectories. To address the above challenges, we propose a Micro-Macro Spatial-Temporal Graph-based Encoder-Decoder (MM-STGED). Specifically, we model each trajectory as a graph to efficiently describe the micro-semantics of trajectory and design a novel message-passing mechanism to learn trajectory representations. Additionally, we extract the macro-semantics of trajectories and further incorporate them into a well-designed graph-based decoder to guide trajectory recovery. Extensive experiments conducted on sparse trajectories with three different sampling intervals that are respectively constructed from two real-world trajectory datasets demonstrate the superiority of our proposed model.

No abstract available

Transportation is a critical sector which is vulnerable to hazardous events like urban floods. Thedisruption in transportation infrastructure affects the whole community in terms of loss of life and property.A resilient transport system is fundamental for resilient communities, and considering the adverse impactsof climate change, it becomes even more important. The Büyükçekmece Basin is a densely populated andeconomically significant area that also includes regions of natural importance requiring preservation. Thus,a flood hazard can affect more people, cause greater monetary losses and have negative impact on theenvironment and nature.The purpose of this research is to present the disruption of road network in Büyükçekmece Basin due toflooding and create an analytical framework. Vulnerability of road systems to rainfall-induced floods isquantified by integrating different models. An integrated framework linking hydrological model andinundation model is established and the relation between flood depth and road network vulnerability isdetermined. Flood hazard maps are created for two different rainfall events with 1 in 50 year and 1 in 100-year return period and flooded road lengths are determined. The flood levels have shown that there are roadsegments that will face moderate and major flooding. In Büyükçekmece Basin where serious floods haveoccurred in the past, measures need to be taken regarding the road network for possible flood risks.

Extreme weather events have significant economic and social impacts, disrupting essential public services like electricity, phone communication, and transportation. This study seeks to understand the performance and resilience of critical infrastructure systems in Houston, Texas, using Hurricane Harvey (2017) as a case study. We surveyed 500 Houston Metropolitan Statistical Area residents after Hurricane Harvey’s landfall about disruption experience in electricity, water, phone/cellphone, internet, public transportation, workplace, and grocery stores. Our household survey data revealed the proportion and duration of disruption in each system. Approximately 70% of respondents reported experiencing electricity outages, while half (51%) had no access to water for up to six days. Two-thirds of surveyed households lacked internet access, and 50% had their phone services disconnected. Additionally, around 71% of respondents were unable to commute to work, and 73% were unable to purchase groceries for their families during this period. We incorporated the household survey responses into the Dynamic Inoperability Input-Output Model (DIIM) to estimate inoperability and economic losses across interconnected sectors. The projected economic loss was estimated to be in the range of $6.7- $9.7 billion when sensitivity analysis is performed with respect to the number of working days. Understanding the resilience of each sector and the inherent interdependencies among them can provide beneficial insight to policymakers for disaster risk management, notably preparedness and recovery planning for future events.

Climate change has led to more frequent flooding of transportation network links because of sea level rise and extreme weather events, compromising mobility. In previous studies, we have identified bridge and tunnel approaches as critical infrastructure; inundation of these links reduces or eliminates capacity, causing queue spillbacks, rerouting, and increased travel times, with effects propagating through the network. This paper builds on previous work by integrating multi-modal demand, agent-based simulation, and strategic infrastructure prioritization under flooding scenarios. The case study examines the New York City transportation network in the U.S., analyzing commuter responses to potential closures of combinations of bridges and tunnels because of flooding, focusing on impacts on mobility and mode choice. Disruption scenarios are designed using historical data and risk maps to reflect realistic flood-induced closures, though without explicit hydrodynamic modeling. Spatial integration of flood data defines precincts, aligning travel demand analysis with vulnerability zones. Results indicate a significant impact on mobility from the closure of the Manhattan Bridge for morning commuters traveling from Brooklyn to Manhattan, while the Carey Tunnel’s closure has a relatively minor effect. Mode choice is significantly affected when public-transit-carrying links are disrupted. These findings demonstrate the critical need to prioritize multimodal link protection in urban resilience planning.

No abstract available

Climate change is increasing the frequency of extreme weather events, posing critical challenges for the resilience of specialized transportation services (STSs) that provide essential mobility for people with disabilities. In the South Korean context, heatwaves, cold spells, and heavy rainfall are particularly relevant because they directly affect health risks, trip demand, and operational reliability, making them central stressors for evaluating STS resilience in Busan. This study examines STS resilience in Busan, South Korea, focusing on three weather stressors: heatwaves, cold spells, and heavy rainfall. Large-scale operational data from the STSs of Busan were analyzed using the 4R (robustness, rapidity, redundancy, and resourcefulness) framework to classify daily service performance into distinct profiles. The analysis revealed that heatwaves coincided with reduced trip demand and shorter waiting times, yet this apparent stability reflected demand suppression rather than genuine robustness. Heavy rainfall produced the most severe disruptions, with longer and more variable waiting times that exacerbated inequities across users. Cold spells were associated with rapid recovery and the preservation of critical trips, although the small number of cases limits broader interpretation. These findings indicate that resilience in STSs is not uniform but event-specific, offering policy insights for strengthening operational stability and promoting equity in accessible transport.

Abstract Understanding changes in mobility patterns during extreme weather is crucial for urban resilience. Existing studies often overlook the transitions between different transportation modes. This study develops a framework that measures the spatiotemporal anomalies of mobilities and builds transition paths across multiple modes to reveal how people adapt to extreme weather. Analyzing four extreme rainfall events in New York City, we find that Citibike riders are most sensitive to rainfall. In the absence of subway disruptions, they tend to switch to the subway. When the subway system is paralyzed, indicating flooding of the system by heavy rainfall, riders shift to For-Hire Vehicles, followed by taxis. Both demonstrate the value of flexible service in urban resilience. The paralysis-prone subway and the uneven distribution of flexible service indicate that the current transit infrastructure lacks coordination and is unprepared for climate change. Recommendations for enhancing urban resilience include upgrading and maintaining the subway system; enhancing inter-transportation-modal coordination; introducing amphibious transportation modes; improving pre-disaster awareness of inland populations; encouraging safety shared ride; and connecting affordable transition paths for underprivileged groups.

Meteorological and climatological trends are surely changing the way urban infrastructure systems need to be operated and maintained. Urban road traffic fluctuates more significantly under the interference of strong wind–rain weather, especially during tropical cyclones. Deep learning-based methods have significantly improved the accuracy of traffic prediction under extreme weather, but their robustness still has much room for improvement. As the frequency of extreme weather events increases due to climate change, accurately predicting spatiotemporal patterns of urban road traffic is crucial for a resilient transportation system. The compounding effects of the hazards, environments, and urban road network determine the spatiotemporal distribution of urban road traffic during an extreme weather event. In this paper, a novel Knowledge-driven Attribute-Augmented Attention Spatiotemporal Graph Convolutional Network (KA3STGCN) framework is proposed to predict urban road traffic under compound hazards. We design a disaster-knowledge attribute-augmented unit to enhance the model’s ability to perceive real-time hazard intensity and road vulnerability. The attribute-augmented unit includes the dynamic hazard attributes and static environment attributes besides the road traffic information. In addition, we improve feature extraction by combining Graph Convolutional Network, Gated Recurrent Unit, and the attention mechanism. A real-world dataset in Shenzhen City, China, was employed to validate the proposed framework. The findings show that the prediction accuracy of traffic speed can be significantly increased by 12.16%~31.67% with disaster information supplemented, and the framework performs robustly on different road vulnerabilities and hazard intensities. The framework can be migrated to other regions and disaster scenarios in order to strengthen city resilience.

In recent years, the evaluation and improvement of road safety efficiency have become increasingly prominent topics in research across various countries, due to the rise in extreme weather conditions. This study examines the factors that impact road safety investment from both government and private transportation perspectives, while also introducing extreme weather as an exogenous variable, and further constructed MPTU-SBM model to systematically assess the road safety efficiency of 30 provinces in China and propose pathways for promotion. The research findings indicate that there are significant regional differences in the efficiency of road safety. Extreme weather has a major impact on road construction and traffic safety. The influence of traffic investment on the efficiency of traffic structure also exhibits spatial heterogeneity across different regions. Furthermore, relevant improvement suggestions have been proposed.

Wheeled machines, such as agricultural tractors, snowplows, and wheeled mobile robots, usually work on icy or snow-covered roads. Therefore, it is very important to study the driving and slip resistance of the tires of these machines. In this paper, we investigate the driving behavior of tires on snow-covered terrain by means of numerical simulations. A high-fidelity snow-covered road model is established, and smoothed particle hydrodynamics (SPH) and the finite element method (FEM) are employed to account for the behaviors of the snow layers and the pavement, respectively. We use the node-to-surface algorithm for the contact interactions between the snow and the pavement. The SPH parameters for the snow are calibrated by means of a triaxial compression experiment. A simplified tire model is established as well, using the FEM, and the effectiveness of the model is demonstrated via comparisons with the experimental data in terms of stiffness. Finally, the tire driving performance on the snow-covered road is simulated, and the influence of the tire surface configuration, external load, inflation pressure, and snowpack compression on the tire traction behaviors is systematically investigated.

Human error is considered to be one of the major causes of crashes, especially in inclement weather. Although many studies have investigated the effect of adverse weather on traffic safety and operations, there is a lack of research into the differences in driving behavior and performance during adverse weather, particularly at a trajectory level. With this research gap in mind, this study presents a novel approach for an in-depth investigation of driver speed selection behavior in adverse weather utilizing trajectory-level data acquired from the SHRP2 Naturalistic Driving Study using a promising association rules data mining technique. The preliminary analysis revealed that drivers reduced their speeds by 3.9% in the presence of light rain, by 10.2% in heavy rain, 15.2% in light snow, 29.8% in heavy snow, 1.8% with distant fog, and 7.4% with near fog. The findings from the association rules mining approach indicated that driving more than 5 mph above the speed limit was closely associated with clear weather as well as young and inexperienced drivers; whereas a reduction in speed to more than 5 mph below the speed limit was closely associated with snowy road surfaces combined with affected visibility. These findings are also in line with the results from the ordered logistic regression, which revealed that drivers were 1.4 times more likely to reduce their speeds in light rain, 1.7 times in heavy rain, 4.3 in light snow, 12.2 in heavy snow, 1.7 with distant fog, and 2.0 with near fog. The findings from this study provide an unprecedented opportunity to develop a Human-in-the-Loop Variable Speed Limit algorithm.

With the continuous development of social economy, traffic safety problems become increasingly prominent. Some cold regions have affected by snow and ice conditions for a long time, accident rate and accident severity increased significantly. Traffic safety problems are particularly serious. So the study of driving behavior in cold regions is an important task for researchers and transportation management authorities. This paper analyzed the influence of snow and ice conditions on freeway traffic, which including drivers, vehicles and road conditions. The traffic operation video data under the normal conditions, as well as under ice and snow conditions, were collected and the driver's lane changing behaviors were analyzed. From the definition of the lane changing process, the distinction of the types, the influencing factors and the kinematics analysis, this paper analyzed the whole lane changing behavior, and explored the specific characteristics of the lane changing frequency. Based on video data can be found that driver's land changing behavior occurs frequently on the freeway, while the frequency of lane changing under snow and ice conditions is significantly higher than that in normal weather, and the proportion of forced lane changing increases, which brings great security risks. This study can be useful for the analysis of the causes of frequent traffic accidents, and play an important role in improving traffic operation efficiency and reducing accidents in cold region.

Abstract This paper describes the authors’ continued efforts toward the development, calibration, validation and application of a large-scale, agent-based model of the Buffalo-Niagara metropolitan area. The model is developed using the TRansportation ANalysis SIMulation System (TRANSIMS), an open-source, agent-based suite of transportation models originally developed by Los Alamos National Lab (LANL). Following the network error-checking, calibration and validation phases of the model development cycle, the model was used to evaluate the impact of significant snow storm events on the performance of surface transportation network. This was done by modifying the behavior of the agents (i.e. the drivers) in the model to reflect more conservative driving behavior and vehicle dynamics limitations (such as maximum acceleration and deceleration) imposed by the impaired road surface condition. The study demonstrates that the development of regional agent-based models is technically feasible, but one that requires significant efforts in terms of network accuracy checking, model calibration and validation. Moreover, it is shown that inclement weather events reduce the ability of transportation networks to handle the travel demand, which in turn underscores the importance of effective travel demand management during such events.

No abstract available

This paper describes a multiple-hypothesis tracking (MHT) formulation of a particular set of situational awareness problems that involve monitoring of spatio-temporal phenomena using ubiquitous sensing. In particular, the focus is on large-scale monitoring (or tracking) applications utilizing volunteer mobile ubiquitous sensors to track the evolution of spatio-temporal ‘targets’ of interest (e.g., tracking snow-fall or ride-quality at a certain stretch of a roadway using vehicles). An efficient framework is developed utilizing MHT as the basis to carry out detection and tracking of evolutionary behavior. The framework is described via an illustrative example on ride-quality monitoring as applied to autonomous driving environments, where vibration and GPS data recorded by voluntarily participating vehicles are utilized for threat detection and tracking.

No abstract available

No abstract available

LiDAR point clouds are significantly impacted by snow in driving scenarios, introducing scattered noise points and phantom objects, thereby compromising the perception capabilities of autonomous driving systems. Current effective methods for removing snow from point clouds largely rely on outlier filters, which mechanically eliminate isolated points. This research proposes a novel translation model for LiDAR point clouds, the ‘L-DIG’ (LiDAR depth images GAN), built upon refined generative adversarial networks (GANs). This model not only has the capacity to reduce snow noise from point clouds, but it also can artificially synthesize snow points onto clear data. The model is trained using depth image representations of point clouds derived from unpaired datasets, complemented by customized loss functions for depth images to ensure scale and structure consistencies. To amplify the efficacy of snow capture, particularly in the region surrounding the ego vehicle, we have developed a pixel-attention discriminator that operates without downsampling convolutional layers. Concurrently, the other discriminator equipped with two-step downsampling convolutional layers has been engineered to effectively handle snow clusters. This dual-discriminator approach ensures robust and comprehensive performance in tackling diverse snow conditions. The proposed model displays a superior ability to capture snow and object features within LiDAR point clouds. A 3D clustering algorithm is employed to adaptively evaluate different levels of snow conditions, including scattered snowfall and snow swirls. Experimental findings demonstrate an evident de-snowing effect, and the ability to synthesize snow effects.

As the prevalence of electric vehicles (EVs) continues to surge, the precise forecasting of charging demands at individual charging stations becomes imperative for effective power distribution management. However, the charging demand of EVs is often related to various factors and exhibits strong randomness. This paper aims to explore the impact of climatic factors on the charging demand of electric vehicles at charging stations, and to study a prediction model based on the attention mechanism of LSTM for predicting the load of charging stations, providing important guidance for the scheduling of electric vehicles. By analysing the load data of a single charging station under different climatic conditions, the paper finds that climatic factors such as the highest temperature of the day, the lowest temperature, and the type of weather significantly affect the charging demand of electric vehicles. Utilizing the aforementioned characteristics, this paper studies a climate feature‐guided charging demand prediction model for a single charging station, which adopts the LSTM architecture and introduces the attention mechanism, incorporating the above‐mentioned important climatic features, and is ultimately able to accurately predict the future charging demand of the charging station. The experimental results show that, compared to other time series forecasting models, this model has significantly improved performance on the dataset tests, with its accuracy ratio (AR) indicator and qualified rate indicator both exceeding 0.85 and 0.95, respectively. This study not only offers a new perspective and method for predicting the demand for electric vehicle charging but also provides support for the development of electric vehicle scheduling.

: Aiming at the problem of increasing the peak-to-valley difference of grid load and the rising cost of user charging caused by the disorderly charging of large-scale electric vehicles, this paper proposes a coordinated charging scheduling strategy for multiple types of electric vehicles based on the degree of urgency of vehicle use. First, considering the range loss characteristics, dynamic time-sharing tariff mechanism, and user incentive policy in the low-temperature environment of northern winter, a differentiated charging model is constructed for four types of vehicles: family cars, official cars, buses, and cabs. Then, we innovatively introduce the urgency parameter of charging demand for multiple types of vehicles and dynamically divide the emergency and non-emergency charging modes according to the difference between the regular charging capacity and the user’s minimum power demand. When the conventional charging capacity is less than the minimum power demand of the vehicle within the specified time, it is the emergency vehicle demand, and this type of vehicle is immediately charged in fast charging mode after connecting to the grid. On the contrary, it is a non-emergency demand, and the vehicle is connected to the grid to choose the appropriate time to charge in conventional charging mode. Finally, by optimizing the objective function to minimize the peak-to-valley difference between the grid and the vehicle owner’s charging cost, and designing the charging continuity constraints to avoid battery damage, it ensures that the vehicle is efficiently dispatched under the premise of meeting the minimum power demand. Simulation results show that the proposed charging strategy can reduce the charging cost of vehicle owners by 26.33%, reduce the peak-to-valley difference rate of the grid by 29.8%, and significantly alleviate the congestion problem during peak load hours, compared with the disordered charging mode, while ensuring that the electric vehicles are not overcharged and meet the electricity demand of vehicle owners. This paper solves the problems of the existing research on the singularity of vehicle models and the lack of environmental adaptability and provides both economic and practical solutions for the cooperative optimization of electric vehicles and power grids in multiple scenarios.

We propose a forecasting technique based on multi-feature data fusion to enhance the accuracy of an electric vehicle (EV) charging station load forecasting deep-learning model. The proposed method uses multi-feature inputs based on observations of historical weather (wind speed, temperature, and humidity) data as multiple inputs to a Long Short-Term Memory (LSTM) model to achieve a robust prediction of charging loads. Weather conditions are significant influencers of the behavior of EV drivers and their driving patterns. These behavioral and driving patterns affect the charging patterns of the drivers. Rather than one prediction (step, model, or variables) made by conventional LSTM models, three charging load (energy demand) predictions of EVs were made depending on different multi-feature inputs. Data fusion was used to combine and optimize the different charging load prediction results. The performance of the final implemented model was evaluated by the mean absolute prediction error of the forecast. The implemented model had a prediction error of 3.29%. This prediction error was lower than initial prediction results by the LSTM model. The numerical results indicate an improvement in the performance of the EV load forecast, indicating that the proposed model could be used to optimize and improve EV load forecasts for electric vehicle charging stations to meet the energy requirements of EVs.

The past decade has witnessed a remarkable surge in adoption of electric vehicles (EVs). The momentum is expected to continue with strong support from governments and industry. Rapid EV adoption will add significant electricity demand, making it critical to plan for and manage EV charging to avoid causing additional stress and non-negligible risks to the already-aging power grid. To help power grid operators understand the impacts of residential EV charging and identify risk factors, this study presents a data-driven charging demand analysis for light-duty vehicles. This study considers two real-world grid service regions in Colorado and merges multiple data sources and state-of-the-art tools that characterize EV adoption projections, vehicle travel patterns, seasonal variations, residential charging accessibility, ambient temperature impact, EV charging behaviors, grid utility customers, vehicle registration, and household-level EV charging demand distribution. We characterize potential residential charging demand in 2030 for two regions within the state of Colorado: Boulder and Aurora regions. We project that EVs will be 26% of the light-duty vehicle population in Boulder and 16% in Aurora areas. Charging demand is characterized for ten power grid feeders (five for each study region). Across the ten feeders, peak total EV charging powers during wintertime range from less than 1 MW to more than 4 MW.