Browse Topic: Traffic management

Items (564)
Find
Research Review on Data Fusion-Based Ship Information Perception Technology2026-99-05717/10/2026
With the development of domestic vessel traffic service (VTS) systems, China has established a comprehensive maritime traffic management infrastructure. Marine sensing equipment, including radar, the automatic identification system (AIS), and electro-optical (EO) systems, provides diverse sources of ship information. In recent years, data fusion technology has attracted increasing attention for its potential to improve the accuracy and completeness of ship perception. This paper introduces key ship information sensing technologies and examines the distinct characteristics of each approach. It then reviews recent advances in three main areas: vision-based ship feature recognition, multi-source data association analysis, and ship motion prediction. Finally, the paper outlines prospective research directions, including the integration of additional data sources, real-time data processing, enhanced data security, and intelligent maritime decision-making.
Zhao, Kui, Song, Zheming, Huang, Yuantao
Path Planning Algorithm for Self-Driving Cars Based on High-Precision Maps and Improved A*2026-99-05887/10/2026
To address the limitations of the traditional A* algorithm in lane-level navigation, we propose an autonomous vehicle path planning algorithm based on high-precision maps and an improved A* algorithm to ensure effective application in complex traffic environments. We construct a hierarchical high-precision map based on the Lanelet2 framework to achieve structured modeling of complex road environments. To address the adaptability issues of the A* algorithm in lane-level navigation, we propose optimization schemes, including heuristic function improvements, path segment division, and target point validity verification, to ensure that vehicles can autonomously change lanes on multi-lane roads. By combining dynamic programming (DP) and quadratic programming (QP), we ensure the safety and smoothness of the path. Simulation results demonstrate that the optimized algorithm enables smooth stopping and starting at traffic lights in structured road environments and autonomous lane changes on multi-lane roads. Compared to using DP alone, QP provides smoother and safer driving paths and exhibits superior obstacle avoidance performance in speed planning. This method effectively ensures the rationality of path planning in complex road environments while strictly adhering to traffic rules, thereby enhancing the safety and reliability of path planning.
Wang, Siyu, Zhou, Rong, Shi, Tian, Xu, Zhen, Zhao, Zhiguo
Traffic Crash Prediction and Causal Disentanglement by Machine Learning2026-99-05527/10/2026
Aiming at the problem of insufficient modeling of spatio-temporal heterogeneity in road traffic accident prediction, a dual task machine learning framework integrating geographical environment, location attributes and time periodicity is proposed. The dataset used in this study was derived from traffic accident records of Nanchang during 2019–2023. Firstly, geographical identifiers are generated by rounding and aggregating latitude and longitude coordinates. At the same time, the location type is processed by a one-hot encoding, so as to carry out spatial clustering analysis of accident hotspots. Compared with the North-South pattern, the contribution of geographical features shows a strong East-West trend. The kernel density heatmap identified Zone A and zone B as dual core high-risk areas. Secondly, the sinusoidal/cosine function is used to encode the time feature circularly, which effectively captures the daily change of the accident. The quantitative analysis of random forest regression model showed that time characteristics accounted for 89.2% of the variance of accident frequency interpretation, significantly exceeding the contribution of geographical factors (10.2%) and location attributes (0.6%). After hyperparameter optimization, the accuracy of XGBoost classifier in predicting serious accidents is 75.97%, and the AUC value is 0.8412, which has strong robustness, and provides reliable support for dynamic risk assessment of traffic management system.
Luo, Jiang, Zhang, Yuxin, Li, Xin, Wu, Ronghai
Advances in Research on Emergency Lane Activation Model for Highways2026-99-05427/10/2026
12
Qin, Fengcai, Chen, Jianqiu, Che, Guoyan, Lou, Benxiao, Wang, Xiang, Ning, Longtang, Zhou, Shixuan, Zhang, Xiyuan, Bao, Chun, Gu, Guobin
Improved Resilience using Tokenized Gamification Engaging Social Eco-systemF-0082-2026-01425/5/2026
Urban air mobility with electric Vertical Take-Off and Landing (eVTOL) aircraft faces critical micro-weather and infrastructure readiness challenges. This paper proposes a novel socio-technical solution: a tokenized gamification platform that crowdsources hyper-local wind and weather data to enhance operational resilience. We outline the safety gap left by traditional aviation weather systems (METAR, AWOS, ASOS) in urban environments, and leverage community engagement to fill it. The proposed system integrates with Unmanned Traffic Management (UTM) and Safety Management Systems (SMS) to validate user-contributed micro-weather observations, incentivize accurate reporting through tokens and skill-level progression, and feed data into AI-driven forecasts. Early proof-of-concept results indicate improved wind hazard detection and robust user participation. By aligning with emerging regulations (FAA, EASA, DGCA) and test frameworks, this crowdsourced micro-weather ecosystem shows potential to uplift eVTOL safety, build public trust, and support city-scale planning for advanced air mobility.
Udipi, Ranga, Raul, Swarab, Esturi, Ankith
Aerodynamic Proximity Effects between the DrivAer and AeroSUV Standard Models, Part 1 – Same-Lane Interactions2026-01-06054/7/2026
In this paper, the effects of aerodynamic interactions on the drag of a longitudinally-arranged two-vehicle system are examined by considering the influence of separation distance, cross winds, vehicle size and shape. Testing was undertaken at 30% scale in a large wind tunnel with road-representative freestream turbulence. Separation distances of 0.5, 1.0, and 2.0 vehicle lengths (L) were examined over a range of yaw angles between ±15°. A highlight of the current study is the characterization of platoon drag-reduction benefits for different sizes and shapes of the lead and follower models, by using a DrivAer model and an Aero-SUV model, each with slant-back (Notchback or Fastback) and square-back (Estateback) variants, providing four distinct model pairings. Drag reduction for the lead model appears to be affected mainly by the size of the follower model, while the follower model shows a much greater sensitivity to shape of the lead model. Larger drag reductions were observed at most distances and yaw angles when the lead model had a slant-back configuration (Notchback or Fastback), with smaller drag reductions observed for lead models with square-back configurations (Estateback). This resulted from the different wake structures and their respective influences on the surface-pressure distributions of the follower model. Thrust sheltering is observed as the dominant cause for increased drag at the shortest separation distance. Most of the data show that the drag reductions for the two-vehicle system were larger when the AeroSUV model followed the DrivAer model. This was due to a combination of the greater proportional drag reduction for the leading DrivAer and to the greater relative weighting of the AeroSUV drag reduction due to its larger reference drag area. Peak system-drag reductions of up to 22% were observed at 0.5L separation, decreasing to 18% at 1.0L and 12% at 2.0L.
McAuliffe, Brian, Ghorbanishohrat, Faegheh
Aerodynamic Proximity Effects between the DrivAer and AeroSUV Standard Models, Part 2 – Adjacent-Lane and Lateral-Offset Interactions2026-01-06064/7/2026
Aerodynamic interactions between two 30%-scale passenger vehicles in close proximity were examined experimentally in a large wind tunnel, with a focus on longitudinal separations up to two vehicle lengths, lateral separations up to one lane width, and combinations thereof. Part 1 of this paper described the longitudinal following (platooning) configurations of these results, while this paper concentrates on adjacent-lane influences and lateral-offset effects when platooning at a single separation distance. Test models were based on the DrivAer and Aero-SUV open-access geometries, each with slant-back (Notchback or Fastback) and square-back (Estateback) variants. This provided four distinct model pairings, not all of which were tested in each positional arrangement. Adjacent-lane results matched the trends from a smaller-scale study in a different wind tunnel using the same geometry pair, with small-but-distinct differences attributed to different blockage ratios in the two wind-tunnel studies. For three specific adjacent-lane arrangements, no significant differences were observed when changing the back variants of either of the models, suggesting that these proximity effects are primarily a function of model size, not shape. Four model pairs were tested with lateral offsets of 0.00, 0.25, 0.50 and 1.00 lane-widths, corresponding to approximately 0, 0.5, 1.0, and 2.0 model widths, at a longitudinal separation distance of 0.5 model lengths. The data suggest that, as crosswinds increase, peak drag reductions from platooning can be maintained by offsetting the vehicles laterally to maintain the follower model in the wake of the lead model, but the effect is sensitive to the shape of the lead vehicle. At 15° yaw angle, a quarter-lane offset (half-width offset) can maintain the system drag reduction at this separation distance.
McAuliffe, Brian, Ghorbanishohrat, Faegheh
Adaptive Eco-Driving Feedback for Battery Electric Vehicles: A Reinforcement Learning Approach Using Context-Aware Policy Optimization2026-01-01654/7/2026
Energy efficiency and range optimization remain critical challenges to the widespread adoption of battery electric vehicles (BEVs). As a result, there is a growing demand for intelligent driver assistance systems that can extend the operating range and reduce range anxiety. This paper presents an adaptive eco-feedback and driver rating system based on proximal policy optimization (PPO) reinforcement learning, designed to support drivers with the target to reduce energy consumption and maximize driving range. The system processes real-time driving data, such as velocity, acceleration and powertrain status. Map data of high quality is used to anticipate traffic events, including but not limited to speed limits, curves, gradients, preceding vehicles and traffic lights. This contextual awareness allows the system to continuously assess driving behavior and provide personalized, context-aware visual feedback alongside a dynamic driving behavior rating. A PPO agent learns optimal feedback strategies through continuous interaction and evaluates the impact of specific guidance actions, such as but not limited to “release accelerator pedal”, “brake” and “recuperate”, on immediate energy efficiency and long-term driver adaptation patterns. Feedback intensity and modality are dynamically tailored to individual driver profiles based on observed reaction patterns and feedback adherence. This approach encourages drivers to prioritize energy efficiency while aiming to minimize cognitive distraction and discomfort. The algorithm is implemented and validated within a driving simulation environment that replicates diverse and realistic conditions. Virtual driving tests conducted in various scenarios, such as congested urban areas, suburban routes, mountain roads and highways demonstrate that the proposed PPO-based eco-driving assistance system can reduce energy losses by about 28% compared to conventional driving behavior.
Stocker, Christoph, Hirz, Mario, Martin, Michael, Kreis, Alexander, Stadler, Severin
Roundabout Dilemma Zone Data Mining and Forecasting with Trajectory Prediction and Graph Neural Networks2026-01-00294/7/2026
Traffic roundabouts, as complex and safety-critical road scenarios, present significant challenges for autonomous vehicles. In particular, predicting and managing dilemma zone (DZ) encounters at roundabout intersections remains a pivotal concern. This paper introduces an AI-driven system that leverages advanced trajectory forecasting to anticipate DZ events, specifically within traffic roundabouts. At the core of our framework is a modular, graph-structured recurrent architecture powered by graph neural networks (GNNs). By modeling agent interactions as a dynamic graph, our approach integrates heterogeneous data sources - including semantic maps - while capturing agent dynamics with high fidelity. This GNN-based forecasting model enables accurate prediction of DZ events and supports safer, data-driven traffic management decisions for both autonomous and human-driven vehicles. We validate our system on a real-world dataset of roundabout intersections, where it achieves high precision with an exceptionally low false positive rate of 0.1. Our work highlights the potential of AI and graph-based deep learning methods for advancing roundabout safety, offering a robust step toward more reliable and intelligent intersection management in the era of autonomous transportation.
Lu, Duo, Satish, Manthan, Farhadi, Mohammad, Chakravarthi, Bharatesh, Yang, Yezhou
Real-time Parameter Optimization for Eco-driving Control in Connected and Automated Vehicles Using Reinforcement Learning2026-01-00414/7/2026
This paper introduces a novel methodology to enhance the energy efficiency of eco-driving controllers in Connected and Automated Vehicles (CAVs) by leveraging reinforcement learning (RL) techniques for real-time parameter optimization. Traditional eco-driving strategies rely on fixed control parameters, which limit adaptability across diverse traffic and road conditions. To address this, we apply continuous action space RL algorithms, specifically Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization (PPO), to dynamically tune four key parameters within a model predictive control framework that is grounded in Pontryagin’s Maximum Principle (PMP). These parameters influence acceleration, braking, cruising, and intersection-approach behaviors, making them critical for achieving optimal eco-driving performance. Our study employs Argonne National Laboratory’s RoadRunner simulator, a Simulink-based environment designed for high-fidelity CAV analysis, incorporating realistic traffic signals, road gradients, and vehicle interactions. RL agents are trained to interpret vehicle states, road attributes, and traffic light information to adjust control parameters in real time. This integration enables the controller to anticipate and respond to dynamic driving scenarios, thereby improving both energy efficiency and operational robustness. Simulation experiments across multiple driving scenarios demonstrate that the RL-enhanced eco-driving controller achieves substantial energy savings without compromising travel time. On average, our approach surpasses a baseline eco-driving controller without RL by 12% and outperforms a high-fidelity human driver model by 24.2% in terms of energy consumption reduction. These results highlight the potential of continuous action space RL to advance real-time eco-driving control in CAVs. Overall, this work provides a pathway toward more intelligent, adaptive, and sustainable vehicle control systems that can accelerate the deployment of energy-efficient mobility solutions.
Zhang, Yaozhong, Ammourah, Rami, Han, Jihun, Moawad, Ayman, Shen, Daliang, Karbowski, Dominik
Quantifying Highway Flow Improvements via CACC-Based Truck Platooning: A Statistical Modeling Approach2026-01-04634/7/2026
The advancement of Cooperative Adaptive Cruise Control (CACC) technology enables vehicle platooning on public roads, offering significant potential to enhance urban mobility, driving safety, and energy efficiency. Among various applications, truck platooning has become a promising strategy to increase highway flow rates by reducing vehicle headways, improving coordination, and optimizing space utilization. This paper presents a quantitative assessment of a CACC-based truck platooning system, focusing on its effectiveness in enhancing highway mobility under varying traffic conditions. A statistical regression model is developed and calibrated using simulations of real-world highway networks to identify key influencing factors and evaluate the resulting improvements in traffic flow. The analysis considers five primary variables: desired platoon speed, platoon size, space headway, percentage of platooning trucks, and non-platoon traffic flow. The study systematically examines the impact of each parameter on overall traffic throughput. Results indicate that truck platooning can increase highway flow rates by up to 200%, particularly under conditions of high truck volumes and larger platoon sizes. Both platoon size and the percentage of platooning trucks show a positive correlation with flow rates, suggesting that greater coordination among vehicles enhances overall mobility. Conversely, higher desired speeds and larger space headways tend to diminish the benefits of platooning by reducing traffic density. Overall, this paper provides a comprehensive quantitative evaluation of the mobility benefits of truck platooning and highlights its potential to significantly improve highway operations. Future work will extend these findings to assess the energy and emission benefits of platooning and to evaluate the performance of large-scale platooning deployment strategies.
Karbasi, Amir Hossein, Wang, Jinghui, Yang, Hao
GTSRB Shield: A Novel Machine Learning Approach to Anomaly Detection in German Traffic Signs Recognition Benchmark (GTSRB) Dataset2026-26-06111/16/2026
The escalating dependence of Autonomous Vehicles on Intelligent Transportation Systems (ITS) has highlighted the imperative for comprehensive security protocols to safeguard such vehicles against cyber threats. Intrusion Detection Systems (IDS’s) are pivotal in ensuring the protection of these systems by detecting and alleviating unauthorized access and nefarious activities. The German Traffic Sign Recognition Benchmark (GTSRB) database, which encompasses an extensive compilation of traffic sign imagery, functions as a vital asset for the advancement of machine learning-based IDS. This research elucidates an intrusion detection system (IDS) that employs machine learning algorithms to scrutinize the GTSRB database. The proposed IDS emphasize the preprocessing of the GTSRB dataset to extricate pertinent features that can be employed for the training of machine learning models. Research also focuses on model development with machine learning algorithms to classify traffic signs and discern anomalies suggestive of potential intrusions. The efficacy of the models is evaluated utilizing accuracy thereby ensuring that the IDS can consistently differentiate between benign and malicious activities. This inquiry contributes to the domain of intelligent transportation systems by establishing a resilient framework in autonomous vehicles for intrusion detection, thus bolstering the security of automated traffic management systems against prospective cyber threats. The results underscore the criticality of incorporating machine learning methodologies in real-time systems to proactively mitigate security vulnerabilities and preserve the integrity of traffic data.
Patil, Kamalesh, Akbar Badusha, A., Jadhav, Savitri, Gunale, Kishanprasad
Integrating V2V and V2I Communication for Enhanced Traffic Safety and Efficiency in Connected Vehicular Networks2026-26-02711/16/2026
This paper is a new approach to improve road safety and traffic flow by combining vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications. The Study is focused on a system that connects vehicles with each other and with traffic light to share real-time data about speed and position. This work is aimed to discuss the methodology adopted for developing a system which predicts and advises the optimal speed for vehicles approaching an intersection. Inspired by the Green Light Optimized Speed Advisory (GLOSA) , the proposed system is designed to help drivers approach traffic signals at speeds that minimize unnecessary stops, reduce delays, and improve traffic efficiency. This paper contains the approach taken, the decision-making algorithm, and the simulation framework built in MATLAB/Simulink to validate the concept under real traffic conditions. Simulation results are presented to demonstrate how the system generates speed recommendations based on vehicle parameters and traffic light states. We have worked on the integration of both V2V and V2I communications, combined with a speed advisory algorithm. This work paves the way for smarter, more responsive traffic management systems and supports the future deployment of connected and autonomous vehicles.
Pinto, Colin Aubrey, Shah, Ravindra, Karle, Ujjwala
Distributed Robust Control of Heterogeneous Vehicular Platoons Based on IDM Following Model under Communication Delay2025-01-735012/31/2025
With the rapid development of automobile industrialization, the traffic environment is becoming increasingly complex, traffic congestion and road accidents are becoming critical, and the importance of Intelligent Transportation System (ITS) is increasingly prominent. In our research, for the problem of cooperative control of heterogeneous intelligent connected vehicle platoons under ITS considering communication delay. The proposed method integrates the nonlinear Intelligent Driver Model (IDM) and a spacing compensation mechanism, aiming to ensure that the platoon maintains structural stability in the presence of communication disturbances, while also enhancing the comfort and safety of following vehicles. Firstly, construct heterogeneous vehicle platoon system based on the third-order vehicle dynamics model, Predecessor-Leader-Following (PLF) communication topology, and the fixed time-distance strategy, while a nonlinear distributed controller integrating the IDM following behavior and the front-vehicle spacing compensation mechanism is designed to enhance the robustness of the system to delay disturbance. Secondly, leveraging the Lyapunov-Krasovskii functional framework in conjunction with the Moon inequality, an LMI-based stability condition is derived to ensure the uniform asymptotic stability of the system. The corresponding maximum admissible communication delay is then determined, followed by a detailed analysis of the system's string stability. Finally, comparative simulations are conducted on the MATLAB/Simulink platform. Simulation results verify that the proposed controller offers enhanced convergence speed, reduced acceleration variability, and improved suppression of spacing errors under communication delay disturbances. Compared to conventional linear controllers, it demonstrates markedly superior control performance and greater practical applicability. This method provides a valuable reference for the robust design and performance optimization of cooperative control systems for heterogeneous vehicle platoons under communication delay conditions.
Ye, Xin, Kang, Zhongping
DMGF-STAN: Spatio-Temporal Attention Network and Dynamic Multi-Graph Fusion for Traffic Flow Prediction2025-99-023812/23/2025
The development of ITS is vital for decreasing traffic congestion and improving traffic scheduling procedures. Traffic prediction is a fundamental component of the development of ITS. Even though a lot of research has been done on modeling intricate spatiotemporal correlations in order to make accurate predictions, traditional methods primarily use predefined graph structures for feature extraction, which leaves out important correlations in the data and leads to limited prediction accuracy. The objective of the DMGF-STAN that we have recently created is to recognize both explicit and latent connections between time and space in traffic flow data that are subjected to various types of alterations. Our framework introduces a dynamic multi-graph expert selection module (DMGE) that combines a multi-graph information aggregation component with a sparse gating network to effectively model complex spatial dependencies. The Dynamic Multi-graph Gating (DMGG) module subsequently integrates global and local spatial feature extraction units-specifically the Adaptive Global Similarity Graph Convolution (AGS-GConv) module and Local Spatio-Temporal Attention Graph Convolution (LSTA-GConv) module-through integration of their outputs via dynamic gating fusion mechanisms. These processed features are then coupled with GRU-based codecs for comprehensive spatio-temporal feature learning, ultimately enabling future traffic state prediction. Our model outperforms the most advanced benchmark approaches in terms of prediction accuracy, according to comparative experiments conducted on real-world traffic datasets. The proposed framework can provide urban traffic management centers with short-term congestion forecasts and support dynamic signal cycle adjustments to reduce average delay during peak hours.
Cheng, Youcai, Bao, Shumei, Ke, Yuhao, Hu, Yongkang
Trajectory Reconstruction at Signalized Intersections in a Vehicle-to-Infrastructure Environment2025-99-021712/23/2025
Vehicle trajectories encapsulate critical spatial-temporal information essential for traffic state estimation, congestion analysis, and operational parameter optimization. In a Vehicle-to-Infrastructure (V2I) environment, connected automated vehicles (CAVs) not only continuously transmit their own real-time trajectory data but also utilize onboard sensors to perceive and estimate the motion states of surrounding regular vehicles (RVs) within a defined communication range. These multi-source data streams, when integrated with fixed infrastructure-based detectors such as speed cameras at intersections, create a robust foundation for reconstructing full-sample vehicle trajectories, thereby addressing data sparsity issues caused by incomplete CAV penetration. Building upon classical car-following (CF) theory, this study introduces a novel trajectory reconstruction framework that fuses CAV-generated trajectories and infrastructure-based speed detection data. The proposed method specifically aims to reconstruct the unobserved trajectories of RVs located between successive CAVs within the same lane, ensuring continuity and accuracy in trajectory estimation. To validate the framework’s effectiveness, extensive SUMO simulations were conducted under different CAV penetration rates (PRs: 5%, 10%, 15%, and 20%) with a controlled traffic flow rate of 1000 veh/h. Key findings indicate that the proposed method maintains stable reconstruction accuracy across all tested penetration rates, with errors remaining within acceptable thresholds. Furthermore, comparative analysis against state-of-the-art CF-based reconstruction approaches reveals substantial improvements in accuracy, achieving reductions of 84.51% (LE), 97.07% (QLE) and 95.55% (TE), respectively. The result highlights the proposed method potential for enhancing real-time traffic state estimation, optimizing signal control strategies, and improving overall traffic management in V2I-enabled urban networks.
Bai, Wei, Fu, Chengxin, Yao, Zhihong
Study on the Evaluation Indicator System of Intelligent Capacity Optimization on Highways2025-99-025312/23/2025
Intelligent capacity optimization of highways could realize intelligent enhancement of traffic capacity by optimizing traffic management, improving traffic efficiency and enhancing system synergy without significantly increasing physical lanes. However, there was a lack of a unified and perfect index system to scientifically evaluate the effectiveness of such projects. This paper analyzed the basic theory, evaluation indicator structure and system, and puts forward seven key evaluation dimensions, which including traffic efficiency enhancement, traffic safety improvement, economic and cost-benefit, environmental impacts, technology application and innovation, system reliability and resilience, and service experience. This paper screened the specific evaluation indexes of the seven dimensions and proposes the hierarchical structure of the index system and the weight determination method. This paper constructed a comprehensive, multi-dimensional evaluation index system for highway smart expansion projects, aiming to provide scientific basis and standardized tools for the planning, decision-making, implementation effect assessment and continuous optimization of highway smart expansion projects.
Che, Xiaolin, Li, Weichen, Zhu, Lili, Li, Xin, Wang, Lin
Design of Intelligent Lighting System for Expressway Tunnel2025-99-023212/23/2025
A smart highway tunnels lighting system based on the technology of cloud platform and Internet of Things(IoTs) has been designed to address the common problems of high energy consumption and low level of intelligence in China's highway tunnel lighting system. The highway tunnel lighting system consists of four layers of architecture: platform management layer, local management layer, middle layer and terminal layer. The system collects real-time brightness, lamp brightness, traffic volume and other data outside the tunnel through various sensors deployed on site, and then uploads the collected data to the main controller through LoRa IoTs. The main controller combines the brightness calculation method of the lighting design rules to control the brightness of the tunnel lighting in real time, achieving real-time adjustment of the brightness of the tunnel LED lights and the brightness outside the tunnel, and realizing a safe and energy-saving lighting effect of "lights on when the car comes, lights on when the car goes, and lights follow the car". The experimental results show that the energy-saving rate of the system has reached about 70%, which has achieved good energy-saving and emission reduction effects, and has significant economic, social, and ecological benefits.
Wang, Juntao, Liu, Jingyang, Liu, Yong, Feng, Xunwei
Basic Diagram Model of Mixed Traffic Flow Considering Buses2025-99-030912/17/2025
The emergence of connected and autonomous vehicle (CAV) technologies has ushered in a new era of mixed traffic flow, where CAVs will coexist with human-driven vehicles (HDVs) for the foreseeable future. To investigate the fundamental relationships among flow, density, and speed in this heterogeneous traffic environment, this study develops a comprehensive analytical framework that explicitly accounts for the impact of bus integration in mixed traffic streams. The study initially identifies vehicle classifications and their respective distribution ratios within heterogeneous traffic streams. A fundamental graphical representation of mixed traffic patterns is established, followed by a comprehensive sensitivity evaluation focusing on free-flow velocity parameters within the proposed framework. Subsequently, a micro-level simulation platform is developed utilizing SUMO software. Research outcomes reveal a favorable link between the percentage of integrated self-driving cars and improvements in traffic flow and congestion measures.
Xiao, Yujie, Chen, Xiufeng, Wang, Meng, Xu, Ying
Optimization of Real-Time Traffic Flow Detection Methods Based on Sensor Networks2025-99-033712/17/2025
As modern society develops rapidly, people’s requests for traffic convenience and traffic safety become greater and greater, and it is essential to eliminate traffic congestion and traffic accident to sustainable development of urban areas. Therefore, this paper brings forward novel solution based on hybrid sensor networks to observe the status of traffic in road networks in order to alleviate traffic jam and prevent traffic accident. With the collection of precise traffic flow information at the time, it realizes traffic flow control at crossroads, gives warning in advance with the congestion or accident. We carried out a bunch of simulation experiments in succession, the main discoveries are as follows. a. The energy consumption is great reduced under the sensor deployment rate between 1:50–1:60 (sensor : vehicles). b.The sampling rates can keep a very high level of precise and efficiency under the appropriate range between 1:50–1:60 (sensor : vehicles).The critical segments of roadways are fitted with the radar sensors to accomplish not only reliable surveillance of traffic congestions but also timeearlies warning of traffic accidents as opposed to relying on the single-sensor network. As reflected in Fig. 17, the heterogeneous sensor network is more robust against sensor errors because of the complementarity effects without relying on individual sensors. The experimental results highlight the potential for hybrid sensing architecture for intelligent transportation systems(ITS) and provides a well-technical basis to alleviate urban traffic jam and improve transportation efficiency.
Wang, Xinhai
2025 International Conference on Transportation Infrastructure and Engineering (ICTIE 2025)2025-99-030012/17/2025
TOC
Real-time Traffic Congestion Prediction: A Novel Online Learning Method with Multi-Head Attention Mechanism and LSTM-Based Integrated Learning2025-99-042012/10/2025
Real-time traffic congestion prediction is essential for proactive traffic management, as it enhances the responsiveness of traffic systems, including route guidance, control, and enforcement. However, the heavy reliance on extensive historical data presents a significant challenge for real-time model updates. To overcome this limitation, this study proposes an advanced online learning framework that integrates a multi-head attention mechanism with LSTM-based ensemble learning. This approach incorporates traffic congestion factors as input features and employs average delay per kilometer as the predictive output. The experimental findings indicate that: 1) the proposed approach successfully enables real-time traffic congestion forecasting, and 2) it demonstrates strong adaptability in dynamic traffic environments.
Fu, Chuanyun, Liu, Jiaming, Lu, Zhaoyou, Wumaierjiang, Ayinigeer, Liu, Huahua, Bai, Wei
Fuel-Efficient Large-Spacing Truck Platooning in Windy Environments Considering Longitudinal and Lateral Motions2025-99-041912/10/2025
Large-spacing truck platooning offers a balance between operational safety and fuel savings. To enhance its performance in windy environments, this study designs a control system integrating both longitudinal and lateral motions. The longitudinal control module regulates the inter-vehicle spacing within a desired range while generating a fuel-optimal torque profile by minimizing unnecessary decelerations and accelerations. The lateral control module ensures lateral stability and maintains alignment between the trucks to achieve the expected fuel savings. A two-truck platoon is simulated with a 3-sec time gap under varying wind conditions, using experimental data from the on-road cooperative truck platooning trials conducted in Canada. The control system effectively remains spacing errors within the preset safety buffer and limits lateral offsets to 0.07 m, ensuring safe and stable platooning in windy environments. Additionally, the smoother speed profiles and reduced lateral offsets help the follower truck achieve fuel savings of up to 4.2%. These results demonstrate the potential of the proposed control system to enable the safe and sustainable deployment of large-spacing truck platooning in real-world windy conditions.
Jiang, Luo, Shahbakhti, Mahdi
Impact of Maximum Platoon Size of Connected and Automated Vehicles on Safety of On-Ramp Mixed Traffic2025-99-045412/10/2025
This study investigates how the maximum platoon size (MaxPS) of Connected and Automated Vehicles (CAVs) influences traffic safety within mixed traffic environment on freeway on-ramps. Built upon the SUMO simulation framework, a mixed traffic flow model involving CAV platoons is developed for on-ramp scenarios. This paper examines traffic conditions under varying on-ramp inflow volumes and evaluates upstream speed fluctuations in the merging area. Safety indicators such as Time Exposed Time-to-Collision (TET) and Time-Integrated time-to-Collision (TIT) are employed to assess overall traffic safety. Additionally, collision types are analyzed. Results indicate that under low on-ramp inflow conditions, a moderate MaxPS with low CAV penetration rates significantly enhances safety, whereas a larger MaxPS is preferable with high penetration rates. Under moderate on-ramp inflow, limiting the CAV MaxPS to 2 reduces conflicts. As on-ramp inflow increases further, a MaxPS of 1 or 2 leads to a lower overall collision risk across different CAV penetration rates. These findings provide insights into optimizing CAV platoon control strategies to enhance safety in mixed traffic environments.
Pan, Gongyu, Huang, Yujie, Xie, Junping
A Cooperative Conflict-Free Algorithm for Autonomous Vehicles at Unsignalized Intersections Based on Incremental Learning Monte Carlo Tree Search2025-99-043412/10/2025
The paper examines how connected automated vehicles (CAVs) can navigate unsignalized intersections—especially those where major roads differ significantly from minor roads. The proposed method uses an improved incremental learning Monte Carlo Tree Search to quickly determine an optimal passing order for vehicles, adjusting in real time based on road conditions and vehicle states. Numerical experiments demonstrate that this approach achieves conflict-free, real-time cooperative, reducing average delays significantly compared to traditional traffic signal control. Compared to fully-actuated signal control, the proposed method achieves average delay reductions of 19.92s, 16.46s, and 15.47s for CAVs across varying demand patterns. The practical application of this research lies in its potential to enhance traffic efficiency in urban areas by replacing traditional signal-based control with intelligent, autonomous intersection management. This could lead to reduced congestion, lower fuel consumption, and improved traffic safety, making it particularly valuable for smart city initiatives and future CAV-dominated transportation systems.
Xue, Yongjie, Gao, Feng, Feng, Qiang, Cui, Shaohua
CAV Optimal Scheduling Method for Unsignalized Intersections Based on Rolling Ant Colony Algorithm2025-99-042812/10/2025
With the development of intelligent networking technology and autonomous driving technology, how to efficiently and safely schedule intelligent networked autonomous vehicles at signalless intersections has become a research hotspot in traffic management. Based on this, this article first designs an objective function that considers both intersection traffic efficiency and intersection traffic safety, taking into account constraints such as safe distance, speed, acceleration, etc., and constructs a signal free intersection CAV traffic scheduling model. On this basis, a model solving algorithm based on rolling ant colony algorithm is proposed. Simulation experiments show that compared with typical signal control methods, this method can significantly improve intersection traffic efficiency and reduce the number of conflicts.
Zhao, Yingjie, Liu, Xiaoming, Ma, Zechao, Wang, Yuanrong
Minimum Requirements to Support Traffic Signal Priority and Preemption™ SET FileJ2945/BS_202511 (Current)11/20/2025
Included in this set are the SAE J2945/B Standard which specifies the over-the-air (OTA) interface between connected vehicles (CVs) and connected intersections (CIs) to support traffic signal priority and preemption (TSPP) applications. It specifies the use of updated revisions of the SAE J2735 Signal Request Message (SRM) and Signal Status Message (SSM) and the use of a Wireless Access in Vehicular Environments (WAVE) Service Advertisement (WSA) to advertise support for TSPP at a CI. Included are a concept of operations, requirements, design, and the Abstract Syntax Notation One (ASN.1) message format, data frame, and data element definitions. Also included is the Abstract Syntax Notation One (ASN.1) file precisely defines the structure of the data used to implement applications conformant to the SAE J2945/B Standard. Using this ASN.1 specification, a compiler tool can be used to produce encodings to enable applications to easily encode and decode the Signal Request Message (SRM) and Signal Status Message (SSM) messages, along with the Wireless Access in Vehicular Environments (WAVE) Service Advertisement (WSA) application data, defined in SAE J2945/B defined in SAE J2945/B. Included in the ASN.1 file is the complete SAE J2735 ASN.1 along with the updates to the ASN.1 for the SRM and SSM defined in SAE J2945/B.
Connected Transportation Interoperability Committee
Minimum Requirements to Support Traffic Signal Priority and Preemption™ ASN FileJ2945/BA_202511 (Current)11/20/2025
This Abstract Syntax Notation One (ASN.1) file precisely defines the structure of the data used to implement applications conformant to the SAE J2945/B Standard. Using this ASN.1 specification, a compiler tool can be used to produce encodings to enable applications to easily encode and decode the Signal Request Message (SRM) and Signal Status Message (SSM) messages, along with the Wireless Access in Vehicular Environments (WAVE) Service Advertisement (WSA) application data, defined in SAE J2945/B. Included in the ASN.1 file is the complete SAE J2735 ASN.1 along with the updates to the ASN.1 for the SRM and SSM defined in SAE J2945/B.
Connected Transportation Interoperability Committee
Minimum Requirements to Support Traffic Signal Priority and PreemptionJ2945/B_202511 (Current)11/20/2025
When identifying the content of this report, one of the goals was that it supports a nationally interoperable method for connected vehicles (CVs) to make traffic signal priority and/or preemption (TSPP) requests of connected intersections (CIs) that support priority and/or preemption services. Given that, this report specifies the over-the-air (OTA) interface between CVs and CIs to support TSPP applications using updated revisions of the SAE J2735 Signal Request Message (SRM) and Signal Status Message (SSM) and the use of a Wireless Access in Vehicular Environments (WAVE) Service Advertisement (WSA) to advertise support for TSPP at a CI. Included are a concept of operations, requirements, design, and message structure definitions developed using a detailed systems engineering process.
Connected Transportation Interoperability Committee
Research on Urban Traffic Congestion Optimisation Strategies Based on V2X Technology2025-99-011711/11/2025
This study investigates urban traffic congestion optimisation strategies based on V2X technology. V2X technology (Vehicles and Internet of Everything) aims to alleviate urban traffic congestion, improve access efficiency, and reduce tailpipe emissions through real-time collection and fusion of traffic data to optimise traffic signal control and path planning. The efficacy of the optimisation strategies under different V2X penetration rates is evaluated by conducting multi-factor orthogonal experiments in different typical congestion scenarios. The experimental results show that the V2X-based signal optimisation, path induction, and event response combination strategies exhibit significant optimisation effects in all three scenarios: node bottleneck, corridor congestion, and event induction. Under the condition of 100% penetration, the combined strategy reduces delay by 41.9% in the node bottleneck scenario, improves accessibility by 28.1% in the corridor congestion scenario, and improves accessibility by 52.0% in the event-induced scenario. In addition, the experiments analysed the correlation between response rate and delay and showed that for every 10% increase in response rate, delay could be reduced by approximately 8.4 seconds. These findings suggest that V2X technology can significantly improve the operational efficiency of urban traffic and provide new solutions for traffic management.
Xi, Chaohu, Li, Jiasheng, Qu, Fengzhen, Liu, Hongjun, Liu, Xiaorui, Wang, Chunpeng
Multimodal Intelligent Transportation Data Fusion and Application Methods in Heavy-Duty Vehicles2025-99-013911/11/2025
Heavy-duty commercial vehicles (HDCVs) are the key mobile nodes in intelligent transportation systems (ITS). However, their complex operating conditions and the diversity of data sources (such as road conditions, driver behavior, traffic signals, and on-board sensors) present considerable difficulties for accurately estimating the state and perceiving the environment using a single modality of data. This requires effective multi-modal data fusion to enhance the control and decision-making capabilities of HDCVs. This paper addresses this need by proposing a customized multi-modal intelligent transportation data fusion framework for intelligent HDCVs. This paper presents a solution for establishing a multi-modal intelligent transportation data collection platform, including real-scene collection methods and simulation scene collection methods based on the SUMO-MATLAB joint simulation platform. Through three representative case studies, the application methods of multi-modal traffic data are demonstrated: vehicle speed prediction, vehicle power demand prediction, and trajectory planning. The hyperparameter optimization using an enhanced LSTM neural network with the Sparrow Search Algorithm (SSA) is achieved, resulting in more adaptable, safer, and more efficient multi-modal intelligent transportation data applications.
Chen, Zhengxian, Wang, Shaoqi, Jiang, Huiming, Zhou, Fojin, Wang, Mingqiang, Li, Jun
An SVM-Based Approach to Develop a New Control Module for H2-Powered Engine using Driver Behaviour Classification2025-32-004211/3/2025
In the next years, the global hydrogen vehicle market is expected to grow at a very high rate. Consequently, it is necessary for scholars and professionals to study and test specific components in order to rise motor efficiency leveraging the new features of connectivity available in smart roads. In particular, our research is focused on the developement of an engine control module driven by evaluation of usage characteristics (e.g., driving style) and "connected-to-x" scenarios using the standard engine control approach. Moreover, the module proposed enables the implementation of "fast running" models to improve the response of vehicles and make the best possible use of H2-powered engine characteristics. That said, in this paper is proposed a new approach to implement the control module, using Support Vector Machine (SVM) as the machine learning algorithm to detect driving style, and consequently modify the parameters of the engine. We choose SVM because i) it is less prone to overfitting; and ii) SVM memory efficiency enables the design of a low-cost, compact size controller board. The first step of our research, described in this paper, is to test the algorithm proposed and verify its performance using the usual machine learning metrics. An open source dataset has been used for training and testing of our SVM-based algorithm and the promising results achieved are shown. As part of future work, this experimental control module will be installed on an H2-powered motor on test bench to assess its functionality and allow proper tuning.
Mastroianni, Michele, Merola, Simona, Irimescu, Adrian, De Santis, Marco, Esposito, Christian, Aversano, Lerina
Improved Continuum Model with Driver Behaviour Nonlinear Analysis2025-99-001110/17/2025
A macroscopic traffic flow model based on car-following models of aggressive and timid drivers is presented in this study. Utilizing differential equation theory, we derive the types and stability characteristics of equilibrium solutions across diverse scenarios within the model. The incorporation of a viscous component improves the system’s stability. Additionally, a branch analysis is performed on the new model to examine the emergence of Hopf and saddle-node bifurcations. Simulation results confirm that the proposed model accurately reflects intricate nonlinear phenomena in traffic flow. Notably, the numerical solutions obtained through data simulation align closely with analytical predictions. Additionally, our findings highlights the importance of incorporating branch analysis in providing complementary insights to existing traffic flow theories.
Yang, ChunFeng, Yang, ChenXiao, Qi, LinYuan, Shi, LongYu, Tang, Qiang, Tan, LiXiang
Solving the Multi-Objective Vehicle Routing Problem with Parcel Lockers2025-99-005010/17/2025
With the development of e-commerce and urbanization, logistics distribution has become a key challenge in improving traffic management and efficiency. The use of parcel lockers can alleviate delivery pressure, enhance user experience, and reduce costs. This paper investigates the Multi-Objective Vehicle Routing Problem with Parcel Lockers (MOVRPPL), aiming to optimize transportation costs, customer satisfaction, and the number of vehicles to improve resource utilization. Based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II), this paper proposes the NSGA-II-NI algorithm, which incorporates the nearest neighbor crossover algorithm and route optimization to approximate the Pareto optimal solution set. Experiments using the Solomon dataset are conducted, and the performance is evaluated using the Inverted Generational Distance (IGD) and Hypervolume (HV), compared with the state-of-the-art algorithm NSGA-II-HI. The experimental results show that our method achieves a better balance among multiple objectives, significantly reducing transportation costs and the number of vehicles while enhancing customer satisfaction and providing superior solutions.
Liu, Yuxin, Wang, Ying
A Review of Current Research and Future Developments in Field Modelling for Traffic Risk Analysis2025-99-005810/17/2025
Discovering the trend of risk changes and formulating risk prevention and control measures are important links in achieving proactive risk prevention and control. Constructing and analyzing field models can visualize the distribution and change of risks and formulate effective risk prevention and control measures. Based on the current situation and trend of field model research, this paper discusses its application in risk identification, aiming to improve the accuracy of risk avoidance. Firstly, different types of field models are classified, and their respective characteristics and application scenarios are introduced. Secondly, the shortcomings in the development of field models are summarised. Finally, in the field of autonomous driving and intelligent traffic management, it is proposed that the accuracy of the model can be improved by multi-scene data fusion, the dynamic response enhances the efficiency of risk avoidance, and the aspect of risk classification in complex environments to enhance the universality of the model provides new ideas for the further application of the field model in the field of intelligent traffic.
Song, Yulian, Yue, Lihong, Wang, Chunxiao
Road Traffic State Classification Method Based on AutoEncoder and Clustering Algorithm2025-99-004810/17/2025
Urban road traffic state classification is essential for identifying early-stage deterioration and enabling proactive traffic management. This study presents a novel method to accurately assess the traffic state of urban roads while addressing the limitations of existing methods in spatial generalization performance. The approach consists of three key components. First, several indicators are designed to capture the spatial-temporal evolution mechanisms of traffic state, speed freedom, flow saturation, and their variations over time and space. Then, a feature learning module based on an AutoEncoder network is introduced to reduce the dimensionality of the constructed feature set. This enhances feature distinction while mitigating noise effects on classification results. Third, k-means clustering is applied to analyze significant features extracted from the AutoEncoder latent space, categorizing road traffic states into fluent, basic fluent, moderate congested and severe congested. Finally, a road network in Xuancheng, a city in Anhui Province, China, is selected as the test area. The results of road state categorization for both the entire network and single roads are visualized and analyzed, demonstrating the interpretability and practical utility of the approach. The proposed method is also compared with classical k-means clustering, the threshold-based classification, and FCM. To quantify performance, we introduce a traffic state fluctuation rate index, defined as the ratio of state transitions between adjacent time windows. The results show that during the daytime (06:00-20:00), the fluctuation index of the proposed method increases by 13.1%, 22.7%, and 29.4% compared to the classical k-means, threshold-based method, and FCM, respectively. Meanwhile, during the nighttime (20:00-24:00 and 00:00-06:00), the fluctuation index decreases by 12.7%, 22.5%, and 9.0%, aligning more closely with the real changing patterns of daytime and nighttime traffic conditions.
Wang, Xiaocong, Huang, Min, Guo, Xinling, Xie, Jiemin, Zhang, Xiaolan
The Morning Commute Problem with Carpool and Dynamic Toll Scheme2025-99-002210/17/2025
This study extends the bottleneck model to analyze dynamic user equilibrium (UE) in carpooling during the morning peak commute. It is assumed that the carpooling platform offers both traditional human-driven vehicles and novel shared autonomous vehicles. First, we analyze the traffic distribution on a two-lane road. We find that traffic distribution is influenced by the additional cost of carpooling behavior. A corresponding functional relationship is established and visualized. Second, we derive the critical fare threshold for carpooling. Carpooling occurs only when the fare is below this threshold. Third, we obtain the user equilibrium (UE) solution under a specified case, including flow distribution, equilibrium cost, and total number of vehicle. Furthermore, a system-optimal dynamic tolling scheme is proposed to minimize total system cost while maintaining commuter UE. By equating the system marginal cost to the equilibrium cost, we derive the analytical expression for the lane toll function. Numerical experiments demonstrate that the toll scheme effectively reduces the total system cost. We find that when the average occupancy rate of carpooling vehicles exceeds a certain threshold, both the user equilibrium cost and total number of vehicles for commuters are shown to decrease. This indicates that more people will choose carpooling. Additionally, the impact of shared autonomous vehicle (SAV) penetration in carpooling fleets on the overall system is examined. We find that an increase in SAV penetration rate has a positive impact at the system level. This research provides insights into the influence of SAV on traditional carpooling services and proposes traffic management strategies tailored to such scenarios, offering new perspectives for conventional traffic management approaches.
Zheng, XiaoLong, Zhong, RenXin
The Traffic Light Control Strategy at Intersections in Intelligent Transportation Systems2025-99-000910/17/2025
This thesis explores strategies for controlling traffic signals at intersections within the context of ITS., emphasizing the role of DRL in optimizing traffic flow. In recent years, urbanization and the rapid increase in vehicle numbers in China have exacerbated traffic congestion, significantly hindering urban development. This study explores innovative approaches to alleviating traffic congestion, focusing on smart traffic signal systems that adjust according to real-time traffic conditions. The research reviews fundamental concepts in traffic signal control, including traffic flow, signal phases, and signal cycles, and investigates how DRL can dynamically adjust traffic light cycles to optimize intersection performance. The findings suggest that DRL provides an effective method for managing complex and unpredictable traffic environments, as it enables systems to self-learn and continuously refine their strategies based on environmental changes. The adoption of this technology holds the potential to greatly optimize traffic flow, alleviate congestion, and boost the performance of urban transportation systems. The study concludes that signal control strategies based on DRL present a viable approach to tackling the issues associated with growing traffic and urban congestion.
Liu, Junao, Zuo, Tingyou
Balancing Safety, Stability, and Comfort in Semi-Truck Platooning: A Study on Safe and Optimal Headways under Real-World Traffic Conditions2025-99-001710/17/2025
On highways, platoons of semi-trucks are a common phenomenon. By maintaining a small headway, these platoons can effectively reduce air resistance, thereby improving fuel efficiency and reducing carbon emissions. However, this driving mode is also accompanied by many safety and operational risks, such as increased risk of rear-end collisions, reduced driving comfort, and susceptibility to interference from other vehicles outside the platoon. Therefore, behavioral analysis and evaluation of semi-truck platoons naturally formed in real traffic environments are of great significance for improving their driving safety, comfort and stability. This study focuses on the headway characteristics of semi-truck platoons, analyzes their headway distribution, headway gap and braking response behavior, and then proposes a safe headway threshold for emergency braking to effectively reduce the probability of rear-end collisions. In addition, the study also defines an optimal headway range to reduce the possibility of external vehicle insertion, thereby improving the overall stability and driving experience of the platoon. Based on this, this paper constructs a semi-truck platoon model with safety as the core, and verifies it with actual traffic data, revealing the behavioral characteristics of naturally formed semi-truck platoons in terms of safety headways, optimal headways, and platoon distributions. The research results not only provide theoretical support for improving the safety and stable operation of naturally formed truck platoons, but also provide technical reference for the deployment and operation of future connected and automated truck (CAT) platoons in real road environments, helping the freight industry to develop in a more efficient and sustainable direction.
Hu, Xiaoqiang, Cao, Qiang
A Study of the Applicability of Variable Guide Lanes at Signalized Intersections2025-99-001310/17/2025
In the context of China’s rapidly expanding urbanization, there is an increasing trend of car ownership among residents, which has led to a concomitant rise in traffic demand and a worsening of traffic congestion. To address this challenge, Variable Guided Lanes have been proposed as a novel traffic management strategy. This strategy entails the real-time adjustment of lane function, in response to fluctuations in traffic flow, with the objective of enhancing intersection access efficiency. The present study employs the average delay of vehicles in the inlet lane of the intersection as the discriminating index, and the left-turn and straight flow in the inlet lane as the discriminating condition. The study establishes an equal average delay model and delineates a threshold curve to assess the suitability of the lane for the implementation of Variable Guided Lanes. Furthermore, the study investigates whether the characteristics of the variable lanes are altered for the applicability study. The simulation experiment, conducted using Vissim, demonstrates that the implementation of Variable Guide Lanes leads to a 24.6% reduction in the total delay of the east entrance lane and enhances intersection capacity. This outcome substantiates the efficacy of the employed modeling method and the viability of establishing Variable Guide Lanes.
Zhang, Qinan, Zhang, Yongzhong
Investigation of Traffic Oscillation Suppression by Periodic-Configuration Vehicular Platoon12-09-02-001210/11/2025
To mitigate traffic oscillation in mixed traffic flow environments, which reduces road capacity and may lead to traffic accidents, this article innovatively proposes a periodic-configuration vehicular platoon to enhance traffic stability, inspired by the vibration attenuation properties of periodic structures. First, the vehicular platoon model is developed based on the periodic structure principle, and the lumped mass method is applied to derive the platoon spacing transfer matrix. Second, the band gap range is calculated based on the common traffic oscillation frequency by appropriately designing the period parameters in the periodic-configuration vehicular platoon. Additionally, the influence of these period parameters on the band gap range is analyzed. Finally, simulation experiments are conducted to analyze the propagation characteristics of traffic oscillations within the platoon, and the relative position diagrams of vehicles in the platoon are obtained. To validate the effectiveness of the periodic-configuration vehicular platoon in mitigating traffic oscillations, a comparative analysis of traffic oscillation suppression is performed between periodic and non-periodic-configuration platoons. The results indicate that, for a vehicular platoon consisting of twenty vehicles, the proposed periodic-configuration platoon can suppress the propagation of traffic oscillations, and the suppression effect is up to 65%. The periodic-configuration vehicular platoon can adjust control parameters for specific frequencies of traffic oscillations to achieve improved traffic flow.
Yang, Xiujian, Zhuang, Qingyuan, Wang, Shenyi
Communication Network Construction Behaviors for Robotic Convoying2025-01-04339/16/2025
We develop a set of communications-aware behaviors that enable formations of robotic agents to travel through communications-deprived environments while remaining in contact with a central base station. These behaviors enable the agents to operate in environments common in dismounted and search and rescue operations. By operating as a mobile ad-hoc network (MANET), robotic agents can respond to environmental changes and react to the loss of any agent. We demonstrate in simulation and on custom robotic hardware a methodology that constructs a communications network by “peeling-off” individual agents from a formation to act as communication relays. We then present a behavior that reconfigures the team’s network topology to reach different locations within an environment while maintaining communications. Finally, we introduce a recovery behavior that enables agents to reestablish communications if a link in the network is lost. Our hardware trials demonstrate the systems capability to operate in real-world environments.
Noren, Charles, Chaudhary, Sahil, Shirose, Burhanuddin, Vundurthy, Bhaskar, Travers, Matthew
Leveraging Advanced Data Technologies for Smart Traffic ManagementEPR20250085/20/2025
The transportation industry is transforming with the integration of advanced data technologies, edge devices, and artificial intelligence (AI). Intelligent transportation systems (ITS) are pivotal in optimizing traffic flow and safety. Central to this are transportation management centers, which manage transportation systems, traffic flow, and incident responses. Leveraging Advanced Data Technologies for Smart Traffic Management explores emerging trends in transportation data, focusing on data collection, aggregation, and sharing. Effective data management, AI application, and secure data sharing are crucial for optimizing operations. Integrating edge devices with existing systems presents challenges impacting security, cost, and efficiency. Ultimately, AI in transportation offers significant opportunities to predict and manage traffic conditions. AI-driven tools analyze historical data and current conditions to forecast future events. The importance of multidisciplinary approaches and educational programs in leveraging AI for transportation applications are emphasized in this report. Click here to access the full SAE EDGETM Research Report portfolio.
Ercisli, Safak
Edge Artificial Intelligence in Connected, Cooperative and Automated MobilityEPR20250095/20/2025
With many stakeholders involved, and major investments supporting it, the advancements in automated driving (AD) are undoubtedly there. Generally speaking, the motivation for advancing AD is driver convenience and road safety. Regarding the development of AD, original equipment manufacturers, technology start-ups, and AD systems developers have taken different approaches for automated vehicles (AVs). Some manufacturers are on the path toward stand-alone vehicles, mostly relying on onboard sensors and intelligence. On the other hand, the connected, cooperative, and automated mobility (CCAM) approach relies on additional communication and information exchange to ensure safe and secure operation. CCAM holds great potential to improve traffic management, road safety, equity, and convenience. In both approaches, there are increasingly large amounts of data generated and used for AD functions in perception, situational awareness, path prediction, and decision-making. The use of artificial intelligence (AI) is instrumental in processing such data, and in that context, “edge AI” is a more recent type of implementation. Edge AI involves AI algorithms in edge computing devices, which requires hardware operating close to where data is generated. This report explores the potential of edge AI in CCAM. Different perspectives on edge AI for CCAM are explored and definitions drafted. Primary applications are explored, and an outlook on further advancements in applications is presented. The report includes a discussion on the benefits, risks, and challenges related to the use of edge AI in this domain. Major issues such as privacy and cybersecurity are considered, as are misconceptions. Furthermore, potential learning benefits, using experiences gained in other sectors, are introduced. NOTE: SAE Edge Research Reports are intended to identify and illuminate key issues in emerging, but still unsettled, technologies of interest to the mobility industry. The goal is to stimulate discussion and work in the hope of promoting and speeding resolution of identified issues. These reports are not intended to resolve the challenges they identify or close any topic to further scrutiny.
Van Schijndel-de Nooij, Margriet, Beiker, Sven
Letter from the Guest Editors12-08-02-00115/6/2025
Letter from the Guest Editors
Liang, Ci, Törngren, Martin
SAE TOMORROW TODAY - Scalable AV Deployment Starts with Fewer Close Calls135135/1/2025
The hallmark of exceptional autonomous driving technology isn't just how it reacts in a crisis but how it avoids one altogether. That's the vision behind May Mobility: a world where self-driving cars confidently navigate busy intersections, unexpected detours, and pedestrian-filled crosswalks with the instincts of a seasoned human driver. At the core of May Mobility's technology platform is its patented Multi-Policy Decision Making (MPDM) system. This breakthrough technology uses real-time, in-situ AI to interpret data, continuously learning and adapting to new, complex, and unpredictable driving conditions. By learning on the fly--much like a human driver--May Mobility's AVs can be deployed faster and more cost-effectively than traditional systems. To explore how May Mobility is scaling its AV technology, we spoke with Ed Olson, CEO and Founder, about the company's city-wide AV deployments, strategic partnerships with Toyota and NTT, and its entrance into the rideshare market through a new collaboration with Lyft. It's an engaging, behind-the-scenes look at how AI-powered mobility solutions are transforming urban transportation and paving the way for safer, smarter roads. We'd love to hear from you. Share your comments, questions and ideas for future topics and guests to podcast@sae.org. Don't forget to take a moment to follow SAE Tomorrow Today--a podcast where we discuss emerging technology and trends in mobility with the leaders, innovators and strategists making it all happen--and give us a review on your preferred podcasting platform. Follow SAE on LinkedIn, Instagram, Facebook, Twitter, and YouTube. Follow host Grayson Brulte on LinkedIn, Twitter, and Instagram.
Hineman, Marcie
A Novel Model for Short-Term Traffic Flow Prediction on Highways Utilizing Multi-Head Attention–BiLSTM–BiGRU12-09-01-00024/24/2025
With the development of intelligent transportation systems and the increasing demand for transportation, traffic congestion on highways has become more prominent. So accurate short-term traffic flow prediction on these highways is exceedingly crucial. However, because of the complexity, nonlinearity, and randomness of highway traffic flows, short-term prediction of its flows can be difficult to achieve the desired accuracy and robustness. This article presents a novel architectural model that harmoniously fuses bidirectional long–short-term memory (BiLSTM), bidirectional gated recurrent unit (BiGRU), and multi-head attention (MHA) components. Bayesian optimization (BO) is also used to determine the optimal set of hyperparameters. Based on the PeMS04 dataset from California, USA, we evaluated the performance of the proposed model across various prediction intervals and found that it performs best within a 5-min prediction interval. In addition, we have conducted comparison and ablation studies. This not only proves the effectiveness of the BO strategy but also highlights the advantages of the proposed model in improving predictive accuracy. These results indicate that our model can effectively handle the complexity of highway traffic flows and provide more accurate traffic flow predictions, thereby significantly improving the operational efficiency of highway traffic.
Chen, Peng, Wang, Tao, Ma, Changxi, Chen, Jun
Cross-Scene Traffic Target Detection and Recognition Based on Improver YOLOv5s12-09-01-00034/24/2025
Aiming at the problem of insufficient cross-scene detection performance of current traffic target detection and recognition algorithms in automatic driving, we proposed an improved cross-scene traffic target detection and recognition algorithm based on YOLOv5s. First, the loss function CIoU of insufficient penalty term in the YOLOv5s algorithm is adjusted to the more effective EIoU. Then, the context enhancement module (CAM) replaces the original SPPF module to improve feature detection and extraction. Finally, the global attention mechanism GCB is integrated with the traditional C3 module to become a new C3GCB module, which works cooperatively with the CAM module. The improved YOLOv5s algorithm was verified in KITTI, BDD100K, and self-built datasets. The results show that the average accuracy of mAP@0.5 is divided into 95.1%, 72.2%, and 97.5%, respectively, which are 0.6%, 2.1%, and 0.6% higher than that of YOLOv5s. Therefore, it shows that the improved algorithm has better detection and recognition accuracy and better generalization performance in cross-scene.
Ning, Qianjia, Zhang, Huanhuan, Cheng, Kehan
Traffic Flow Prediction Based on the Optuna–CNN–BiLSTM–GBRT Model12-09-01-00014/24/2025
Traffic flow prediction is very important in traffic-related fields, and increasing prediction accuracy is the primary goal of traffic prediction research. This study proposes a new traffic flow prediction method, which uses the CNN–BiLSTM model to extract features from traffic data, further models these features through GBRT, and uses Optuna to tune important hyperparameters of the overall model. The main contribution of this study is to propose a new combination model with better performance. The model integrates two deep learning models that are widely used in this field and creatively uses GBRT to process the output features of the front-end model. On this basis, the optimal hyperparameters and the robustness of the model are deeply explored, providing an effective and feasible solution to the difficult problems in traffic flow prediction. This model is experimentally studied using three different data transformation methods (original data, wavelet transform, Fourier transform). Compared with other models using similar data, the evaluation index performance of this model under wavelet transform is better, RMSE, MAE, MAPE, and R 2 values are 53.879, 40.641, 11.13%, and 0.97, respectively, which are better than other comparison models. The results show that the proposed Optuna–CNN–BiLSTM–GBRT model can significantly improve the accuracy of traffic flow prediction and provide an effective means to solve problems in the field of traffic prediction.
Ma, Changxi, Jin, Renzhe
Robust V2X Cruise Control for Class 8 Trucks in the Presence of Traffic Lights2025-01-80334/1/2025
An implementation of a robust predictive cruise control method for class 8 trucks utilizing V2X communication with connected traffic lights is presented in this work. This method accounts for traffic signal phases with the goal of reducing energy consumption when possible while respecting safety concerns. Tightened constraints are created using a robust model predictive control (RMPC) framework in which constraints are modified so that the safety critical requirements are satisfied even in the presence of disturbances, while requiring only the expected bounds of the disturbances to be provided. In particular, variation in the actuator performance under different conditions presents a unique challenge for this application, which the approach applied in this work is well-suited to handle. The errors resulting from lower-level control and actuator performance are accounted for by treating them as bounded and additive disturbances on the states of the model used in the higher level MPC, and the RMPC method is demonstrated to satisfy constraints in the presence of arbitrary bounded disturbances that can be modeled in this way. Simulation results show that these tightened constraints successfully account for error due to low-level control and actuator performance for class 8 trucks. Furthermore, tests were performed on hardware which show the capability for real-time application.
Ellison, Evan, Ward, Jacob, Brown, Lowell, Bevly, David M.
Items per page:
1 – 50 of 564