Pub Date : 2025-10-03DOI: 10.1109/TITS.2025.3615073
Weihang Pan;Binbin Lin;Yafei Wang;Zhengxu Yu;Xinkui Zhao;Xiaofei He;Jieping Ye
The decision-making process for connected and autonomous vehicles (CAVs) at unsignalized intersections is a critical and challenging problem. Previous methods predominantly concentrate on optimizing passage strategies for individual intersections in isolation. However, they often neglect global traffic conditions and task priorities in closed, multi-intersection transportation scenarios, leading to localized congestion. In this work, we propose a method that aims to optimize the passing order of intersections from a global and long-term perspective to enhance overall transportation efficiency. Specifically, we model the coordination of multiple unsignalized intersections as a multi-agent sequential decision problem and solve it through a two-stage method. In the planning stage, we construct fully connected undirected graphs based on vehicle conflict relationships and use the multi-agent proximal policy optimization (MAPPO) algorithm to learn the global priorities. In the scheduling stage, the local vehicle scheduling is formalized as a multi-objective optimization problem. The learned global priorities are soft constraints, while a hybrid filtered beam search determines safe and efficient CAV passing orders. Extensive offline experiments and online tests on real-world and synthetic datasets demonstrate that our proposed method outperforms state-of-the-art approaches in minimizing congestion and enhancing transportation efficiency.
{"title":"Cooperative Driving at Multiple Unsignalized Intersections in Fully Autonomous Driving Scenarios","authors":"Weihang Pan;Binbin Lin;Yafei Wang;Zhengxu Yu;Xinkui Zhao;Xiaofei He;Jieping Ye","doi":"10.1109/TITS.2025.3615073","DOIUrl":"https://doi.org/10.1109/TITS.2025.3615073","url":null,"abstract":"The decision-making process for connected and autonomous vehicles (CAVs) at unsignalized intersections is a critical and challenging problem. Previous methods predominantly concentrate on optimizing passage strategies for individual intersections in isolation. However, they often neglect global traffic conditions and task priorities in closed, multi-intersection transportation scenarios, leading to localized congestion. In this work, we propose a method that aims to optimize the passing order of intersections from a global and long-term perspective to enhance overall transportation efficiency. Specifically, we model the coordination of multiple unsignalized intersections as a multi-agent sequential decision problem and solve it through a two-stage method. In the planning stage, we construct fully connected undirected graphs based on vehicle conflict relationships and use the multi-agent proximal policy optimization (MAPPO) algorithm to learn the global priorities. In the scheduling stage, the local vehicle scheduling is formalized as a multi-objective optimization problem. The learned global priorities are soft constraints, while a hybrid filtered beam search determines safe and efficient CAV passing orders. Extensive offline experiments and online tests on real-world and synthetic datasets demonstrate that our proposed method outperforms state-of-the-art approaches in minimizing congestion and enhancing transportation efficiency.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23298-23313"},"PeriodicalIF":8.4,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665793","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-02DOI: 10.1109/TITS.2025.3612736
Ruiqi Ren;Peixin Shi;Jinwoo Kim
Detection and repair of pavement distress in time are crucial to maximize functional performance and service life, while minimizing maintenance costs on extensive roadway networks. Manual distress detection is labor intensive and error prone. While deep learning techniques offer unparalleled capabilities for automated and accurate pixel-level pavement distress segmentation, their reliance on extensive manual annotations remains a bottleneck. To address this challenge, we propose an open-ended self-supervised framework enabling flexible integration of various pretext tasks for pavement distress segmentation without manual annotations. We introduce a spatial contexts-informed pretext task that automatically generates pseudo labels by leveraging the highly consistent semantic information inherent across continuous pavement images within localized areas. A multi-line parallel network architecture is then employed, where each line extracts a distinct deep representation aligned with the pseudo-label generation process. These representations are jointly optimized through a shared weight update scheme augmented by momentum encoders to capture long-range dependencies. A vision transformer processes the input images during inference, utilizing self-attention to highlight distressed regions based on the learned representations for precise segmentation. Extensive evaluations validate the performance of our framework, outperforming state-of-the-art self-supervised methods by 0.075 mIoU on average, while remarkably surpassing weakly supervised techniques requiring manual image-level annotations. These results are far more promising given that our self-supervised approach avoids human labeling costs, striking a trade-off between model effectiveness and annotation efficiency for large-scale deployments. It helps transportation agencies to realize timely, proactive infrastructure maintenance through scalable, accurate distress monitoring over extensive road networks.
{"title":"A Spatial Contexts-Informed Self-Supervised Learning Approach for Pavement Distress Segmentation","authors":"Ruiqi Ren;Peixin Shi;Jinwoo Kim","doi":"10.1109/TITS.2025.3612736","DOIUrl":"https://doi.org/10.1109/TITS.2025.3612736","url":null,"abstract":"Detection and repair of pavement distress in time are crucial to maximize functional performance and service life, while minimizing maintenance costs on extensive roadway networks. Manual distress detection is labor intensive and error prone. While deep learning techniques offer unparalleled capabilities for automated and accurate pixel-level pavement distress segmentation, their reliance on extensive manual annotations remains a bottleneck. To address this challenge, we propose an open-ended self-supervised framework enabling flexible integration of various pretext tasks for pavement distress segmentation without manual annotations. We introduce a spatial contexts-informed pretext task that automatically generates pseudo labels by leveraging the highly consistent semantic information inherent across continuous pavement images within localized areas. A multi-line parallel network architecture is then employed, where each line extracts a distinct deep representation aligned with the pseudo-label generation process. These representations are jointly optimized through a shared weight update scheme augmented by momentum encoders to capture long-range dependencies. A vision transformer processes the input images during inference, utilizing self-attention to highlight distressed regions based on the learned representations for precise segmentation. Extensive evaluations validate the performance of our framework, outperforming state-of-the-art self-supervised methods by 0.075 mIoU on average, while remarkably surpassing weakly supervised techniques requiring manual image-level annotations. These results are far more promising given that our self-supervised approach avoids human labeling costs, striking a trade-off between model effectiveness and annotation efficiency for large-scale deployments. It helps transportation agencies to realize timely, proactive infrastructure maintenance through scalable, accurate distress monitoring over extensive road networks.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23419-23430"},"PeriodicalIF":8.4,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Vehicle trajectory prediction is gaining significant attention from academia and industry because of its vital role in autonomous driving. However, current theories face two challenges. First, they generally underperform when faced with longer historical trajectory inputs, especially in large-scale scenarios. Second, they usually ignore the temporal continuity of the target vehicle itself. To address these issues, our study proposed a novel extended temporal sequence fusion attention (ETSFA) network. This network can fully capture the information from the historical trajectory and the dynamic influences of adjacent agents. In addition, the novel dual-channel decoupled model can precisely characterize the intricate spatiotemporal interplay of on-road vehicles. Specifically, the proposed network consists of two main parts. For temporal analysis, the linear inference network (LIN) is reparameterized into complex diagonal forms at the state–space model (SSM) layer to express the linear recurrence capability, thus effectively mining the long-term historical trajectory temporal features of the target vehicle. For spatial analysis, an advanced spatial perception module (SPM) based on graph attention networks (GATs) is proposed to aggregate vehicle and intervehicle interaction features. In addition, a spatial inference module (SIM) based on a convolutional linear inference unit (CONVLIN) is customized for spatiotemporal graph features. Finally, the proposed ETSFA is trained and validated across diverse public datasets, including HighD and NGSIM, demonstrating a marked improvement in the prediction accuracy of the proposed ETSFA over existing methods.
{"title":"Enhanced Multi-Vehicle Trajectory Prediction via an Extended Temporal Sequence Fusion Attention Network","authors":"Dengyu Xiao;Yu Sun;Huayan Pu;Weijia Jia;Mingliang Zhou","doi":"10.1109/TITS.2025.3612201","DOIUrl":"https://doi.org/10.1109/TITS.2025.3612201","url":null,"abstract":"Vehicle trajectory prediction is gaining significant attention from academia and industry because of its vital role in autonomous driving. However, current theories face two challenges. First, they generally underperform when faced with longer historical trajectory inputs, especially in large-scale scenarios. Second, they usually ignore the temporal continuity of the target vehicle itself. To address these issues, our study proposed a novel extended temporal sequence fusion attention (ETSFA) network. This network can fully capture the information from the historical trajectory and the dynamic influences of adjacent agents. In addition, the novel dual-channel decoupled model can precisely characterize the intricate spatiotemporal interplay of on-road vehicles. Specifically, the proposed network consists of two main parts. For temporal analysis, the linear inference network (LIN) is reparameterized into complex diagonal forms at the state–space model (SSM) layer to express the linear recurrence capability, thus effectively mining the long-term historical trajectory temporal features of the target vehicle. For spatial analysis, an advanced spatial perception module (SPM) based on graph attention networks (GATs) is proposed to aggregate vehicle and intervehicle interaction features. In addition, a spatial inference module (SIM) based on a convolutional linear inference unit (CONVLIN) is customized for spatiotemporal graph features. Finally, the proposed ETSFA is trained and validated across diverse public datasets, including HighD and NGSIM, demonstrating a marked improvement in the prediction accuracy of the proposed ETSFA over existing methods.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23245-23256"},"PeriodicalIF":8.4,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665776","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-29DOI: 10.1109/TITS.2025.3611411
Yuan Cao;Shuyi He;Feng Wang;Shuai Su;Yongkui Sun
The rail surface defects directly impact the safety and efficiency of heavy-haul train operations. Timely assessment of these defects is crucial for informed maintenance decisions, with precise defect detection at its core. In recent years, the accumulation of extensive rail inspection images has led to the application of numerous computer vision-based methods for pixel-level detection of rail surface defects. However, given the constraint of a limited number of labeled defect samples, ensuring the generalization and robustness of existing methods remains challenging, particularly across varying track conditions and complex heavy-haul scenarios. Thus, this paper introduces a Segment-Anything-Model (SAM)-enhanced method for the detection of rail surface defects. First, a shadow-detection-based algorithm is developed to extract the rail regions and mitigate background interference. Then a student-teacher-Simi-network (S-T-Simi)-based unsupervised method is designed to generate prompt information for SAM. Utilizing this prompt information, we develop a task-specified SAM for precise rail defect detection. Finally, comprehensive validation is performed using inspection data collected from diverse heavy-haul tracks. Experimental results indicate that the proposed method achieves highly accurate segmentation of rail defects.
{"title":"A Large-Model-Enhanced Method for Rail Surface Defect Detection in Heavy-Haul Railway","authors":"Yuan Cao;Shuyi He;Feng Wang;Shuai Su;Yongkui Sun","doi":"10.1109/TITS.2025.3611411","DOIUrl":"https://doi.org/10.1109/TITS.2025.3611411","url":null,"abstract":"The rail surface defects directly impact the safety and efficiency of heavy-haul train operations. Timely assessment of these defects is crucial for informed maintenance decisions, with precise defect detection at its core. In recent years, the accumulation of extensive rail inspection images has led to the application of numerous computer vision-based methods for pixel-level detection of rail surface defects. However, given the constraint of a limited number of labeled defect samples, ensuring the generalization and robustness of existing methods remains challenging, particularly across varying track conditions and complex heavy-haul scenarios. Thus, this paper introduces a Segment-Anything-Model (SAM)-enhanced method for the detection of rail surface defects. First, a shadow-detection-based algorithm is developed to extract the rail regions and mitigate background interference. Then a student-teacher-Simi-network (S-T-Simi)-based unsupervised method is designed to generate prompt information for SAM. Utilizing this prompt information, we develop a task-specified SAM for precise rail defect detection. Finally, comprehensive validation is performed using inspection data collected from diverse heavy-haul tracks. Experimental results indicate that the proposed method achieves highly accurate segmentation of rail defects.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23328-23341"},"PeriodicalIF":8.4,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665721","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-29DOI: 10.1109/TITS.2025.3607003
Xinyu Wang;Chengchuan An;Jingxin Xia;Zhenbo Lu
Trajectory prediction is crucial for autonomous vehicle (AV) trajectory planning. The deep learning based trajectory prediction models are easily manipulated by cyber attack such as adversarial attack or confidential information tampering. Current research in adversarial attack typically relies on vehicle physical motion boundaries to conduct linear search, which limits the diversity of samples and covers up the vulnerabilities of model. Moreover, the reckless driving behaviors underlying the generated trajectory samples can be easily detected and smoothed. In this study, a dual constraint optimization framework for adversarial attack is developed. The proposed framework integrates hard constraint of physical boundary with soft constraint of driving risk map to simulate the actual vehicles interaction. Subsequently, Stochastic Gradient Descent (SGD) incorporates Hard-Soft constraint to increase the search space of local optimal solution. The high-precision vehicle trajectory data (sampling interval 0.1s) from the Next Generation Simulation (NGSIM) dataset supports microscopic traffic flow analysis and is used for validating our methods. The vulnerability of the prediction model is revealed from number of attack frames and input features. Results show that our proposed method increases the Average Displacement Errors (ADE) by 42.04% and Final Displacement Error (FDE) by 24.19% compared to the state-of-the-art method.
{"title":"Effective Adversarial Attack Approach to Assess the Vulnerability of Autonomous Vehicle Trajectory Prediction Models","authors":"Xinyu Wang;Chengchuan An;Jingxin Xia;Zhenbo Lu","doi":"10.1109/TITS.2025.3607003","DOIUrl":"https://doi.org/10.1109/TITS.2025.3607003","url":null,"abstract":"Trajectory prediction is crucial for autonomous vehicle (AV) trajectory planning. The deep learning based trajectory prediction models are easily manipulated by cyber attack such as adversarial attack or confidential information tampering. Current research in adversarial attack typically relies on vehicle physical motion boundaries to conduct linear search, which limits the diversity of samples and covers up the vulnerabilities of model. Moreover, the reckless driving behaviors underlying the generated trajectory samples can be easily detected and smoothed. In this study, a dual constraint optimization framework for adversarial attack is developed. The proposed framework integrates hard constraint of physical boundary with soft constraint of driving risk map to simulate the actual vehicles interaction. Subsequently, Stochastic Gradient Descent (SGD) incorporates Hard-Soft constraint to increase the search space of local optimal solution. The high-precision vehicle trajectory data (sampling interval 0.1s) from the Next Generation Simulation (NGSIM) dataset supports microscopic traffic flow analysis and is used for validating our methods. The vulnerability of the prediction model is revealed from number of attack frames and input features. Results show that our proposed method increases the Average Displacement Errors (ADE) by 42.04% and Final Displacement Error (FDE) by 24.19% compared to the state-of-the-art method.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23231-23244"},"PeriodicalIF":8.4,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665744","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-25DOI: 10.1109/TITS.2025.3610660
Arun Sekar Rajasekaran;Mohammad S. Obaidat;Azees Maria;Kalyan Sundar Kola;Ashok Kumar Das;Youngho Park
Ensuring the safe transfer of information plays an important role in the development of Industry 4.0. Robust authentication and security frameworks are required to establish confidence among vehicles, provide reliable data flow, and improve the overall safety of vehicle platoons. This is crucial for preventing cyber-attacks that could cause accidents or disrupt the synchronized movement of vehicle platoons. Initially, in this study, a novel privacy-preserving mechanism based on an authentication code and cipher test is suggested. Second, an effective authentication system is proposed for vehicle platoons. Third, a novel tree-based group key-sharing mechanism for the exchange of information between vehicle users is proposed. The proposed scheme also supports the sharing of the same group key for entities in Industry 4.0. Finally, a planar array consisting of four elements was specifically built for use in vehicular ad hoc network (VANET) applications operating inside the Dedicated Short-range Communications (DSRC) (802.11p) band to prove the efficacy in terms of practical implementation. To assess the security level of the suggested authentication scheme, both formal and informal analyses were conducted. Finally, the performance of the suggested protocol is evaluated in terms of computational and communication overheads. Moreover, the designed antenna provides complete impedance bandwidth coverage, good gain, and minimal cross-polarization suppression at the optimum frequency of operation in the C band.
{"title":"TGKAV: Tree-Based Group Key Agreement Scheme With Practical Antenna Implementation for Vehicle Platoon","authors":"Arun Sekar Rajasekaran;Mohammad S. Obaidat;Azees Maria;Kalyan Sundar Kola;Ashok Kumar Das;Youngho Park","doi":"10.1109/TITS.2025.3610660","DOIUrl":"https://doi.org/10.1109/TITS.2025.3610660","url":null,"abstract":"Ensuring the safe transfer of information plays an important role in the development of Industry 4.0. Robust authentication and security frameworks are required to establish confidence among vehicles, provide reliable data flow, and improve the overall safety of vehicle platoons. This is crucial for preventing cyber-attacks that could cause accidents or disrupt the synchronized movement of vehicle platoons. Initially, in this study, a novel privacy-preserving mechanism based on an authentication code and cipher test is suggested. Second, an effective authentication system is proposed for vehicle platoons. Third, a novel tree-based group key-sharing mechanism for the exchange of information between vehicle users is proposed. The proposed scheme also supports the sharing of the same group key for entities in Industry 4.0. Finally, a planar array consisting of four elements was specifically built for use in vehicular ad hoc network (VANET) applications operating inside the Dedicated Short-range Communications (DSRC) (802.11p) band to prove the efficacy in terms of practical implementation. To assess the security level of the suggested authentication scheme, both formal and informal analyses were conducted. Finally, the performance of the suggested protocol is evaluated in terms of computational and communication overheads. Moreover, the designed antenna provides complete impedance bandwidth coverage, good gain, and minimal cross-polarization suppression at the optimum frequency of operation in the C band.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23342-23357"},"PeriodicalIF":8.4,"publicationDate":"2025-09-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-24DOI: 10.1109/TITS.2025.3604609
Simona Sacone
Summary form only: Abstracts of articles presented in this issue of the publication.
仅以摘要形式提供：本刊发表的文章摘要。
{"title":"Scanning the Issue","authors":"Simona Sacone","doi":"10.1109/TITS.2025.3604609","DOIUrl":"https://doi.org/10.1109/TITS.2025.3604609","url":null,"abstract":"Summary form only: Abstracts of articles presented in this issue of the publication.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 9","pages":"12780-12797"},"PeriodicalIF":8.4,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11178165","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145315492","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-24DOI: 10.1109/TITS.2025.3604346
{"title":"IEEE Intelligent Transportation Systems Society Information","authors":"","doi":"10.1109/TITS.2025.3604346","DOIUrl":"https://doi.org/10.1109/TITS.2025.3604346","url":null,"abstract":"","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 9","pages":"C3-C3"},"PeriodicalIF":8.4,"publicationDate":"2025-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11178164","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145128459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-22DOI: 10.1109/TITS.2025.3608730
Siwen Yang;Yunwen Xu;Dewei Li
This study proposes a mixed traffic flow control structure, utilizing multi-lane fleets composed of a small portion of connected and autonomous vehicles (CAVs) in mixed flow as mobile actuators. Multiple fleets formed within the mixed traffic flow feature a novel structure and a valve-like function that regulates both traffic volume and speed, referred to as “valve fleet”. Specifically, the travel speed and structural spacing of valve fleets are controllable parameters, which can regulate the surrounding traffic speed and the flow through the fleet based on the downstream traffic state to decongest bottlenecks. Then, a control-oriented mobile cell transmission model (MCTM) is developed to characterize the macroscopic traffic dynamics with the presence and influence of valve fleets and bottleneck areas on freeways. Moreover, a hierarchical framework for mixed traffic flow regulation is designed, where the upper-level traffic optimization model dynamically determines all fleets’ parameters in a rolling-horizon fashion to minimize total travel time and suppress local congestion. The decentralized lower-level fleet controller adopts model predictive control (MPC) to coordinate CAVs’ motions and handles interactions between CAVs and human-driven vehicles (HDVs). To evaluate the proposed method, we conduct microscopic experiments in the SUMO simulator to implement the proposed traffic control method in realistic traffic environments. The comprehensive comparison results demonstrate the proposed method’s superiority in enhancing traffic efficiency and alleviating congestion at freeway bottleneck segments.
{"title":"Valve Fleets: A Novel Control Method for Mixed Traffic Flow Regulation With Applications in Bottleneck Segments","authors":"Siwen Yang;Yunwen Xu;Dewei Li","doi":"10.1109/TITS.2025.3608730","DOIUrl":"https://doi.org/10.1109/TITS.2025.3608730","url":null,"abstract":"This study proposes a mixed traffic flow control structure, utilizing multi-lane fleets composed of a small portion of connected and autonomous vehicles (CAVs) in mixed flow as mobile actuators. Multiple fleets formed within the mixed traffic flow feature a novel structure and a valve-like function that regulates both traffic volume and speed, referred to as “valve fleet”. Specifically, the travel speed and structural spacing of valve fleets are controllable parameters, which can regulate the surrounding traffic speed and the flow through the fleet based on the downstream traffic state to decongest bottlenecks. Then, a control-oriented mobile cell transmission model (MCTM) is developed to characterize the macroscopic traffic dynamics with the presence and influence of valve fleets and bottleneck areas on freeways. Moreover, a hierarchical framework for mixed traffic flow regulation is designed, where the upper-level traffic optimization model dynamically determines all fleets’ parameters in a rolling-horizon fashion to minimize total travel time and suppress local congestion. The decentralized lower-level fleet controller adopts model predictive control (MPC) to coordinate CAVs’ motions and handles interactions between CAVs and human-driven vehicles (HDVs). To evaluate the proposed method, we conduct microscopic experiments in the SUMO simulator to implement the proposed traffic control method in realistic traffic environments. The comprehensive comparison results demonstrate the proposed method’s superiority in enhancing traffic efficiency and alleviating congestion at freeway bottleneck segments.","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 12","pages":"23215-23230"},"PeriodicalIF":8.4,"publicationDate":"2025-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145665768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The development of surround-view cameras is crucial for the advancement of autonomous driving. Utilizing depth information and image features to simulate LiDAR bird’s-eye-view (BEV) features can accomplish efficient 3D object detection tasks. Existing dense BEV generation methods heavily rely on the use of depth features, however, the suboptimal exploitation of these features often results in ambiguity in object location and feature representation during the BEV generation process. To address this, we have designed a hybrid encode and distillation method to enhance 3D object detection performance, termed HybridBEV. Initially, we designed the HybridEncode module, which employs a resampling strategy of depth features in voxel space to obtain BEV features that more accurately reflect the distribution of objects. Subsequently, we introduced multiple distillation methods to supervise the network’s voxel features and BEV feature representations, assisting the student network in learning critical features from the teacher model and ensuring that BEV features can more distinctly represent object distribution. Furthermore, during network training, we loaded pre-trained weights from the teacher network to guide network optimization and accelerate training. Extensive experiments on the nuScenes benchmark demonstrate that HybridBEV can effectively improve the performance of the student network and outperform previous state-of-the-art methods based on surround-view cameras. The code will be published at https://github.com/wjyxx/HybridBEV
{"title":"HybridBEV: Hybrid Encode and Distillation for Improved BEV 3D Object Detection","authors":"Junyin Wang;Chenghu Du;Huikai Liu;Zhenchang Xia;Bingyi Liu;Shengwu Xiong","doi":"10.1109/TITS.2025.3599015","DOIUrl":"https://doi.org/10.1109/TITS.2025.3599015","url":null,"abstract":"The development of surround-view cameras is crucial for the advancement of autonomous driving. Utilizing depth information and image features to simulate LiDAR bird’s-eye-view (BEV) features can accomplish efficient 3D object detection tasks. Existing dense BEV generation methods heavily rely on the use of depth features, however, the suboptimal exploitation of these features often results in ambiguity in object location and feature representation during the BEV generation process. To address this, we have designed a hybrid encode and distillation method to enhance 3D object detection performance, termed HybridBEV. Initially, we designed the HybridEncode module, which employs a resampling strategy of depth features in voxel space to obtain BEV features that more accurately reflect the distribution of objects. Subsequently, we introduced multiple distillation methods to supervise the network’s voxel features and BEV feature representations, assisting the student network in learning critical features from the teacher model and ensuring that BEV features can more distinctly represent object distribution. Furthermore, during network training, we loaded pre-trained weights from the teacher network to guide network optimization and accelerate training. Extensive experiments on the nuScenes benchmark demonstrate that HybridBEV can effectively improve the performance of the student network and outperform previous state-of-the-art methods based on surround-view cameras. The code will be published at <uri>https://github.com/wjyxx/HybridBEV</uri>","PeriodicalId":13416,"journal":{"name":"IEEE Transactions on Intelligent Transportation Systems","volume":"26 11","pages":"21257-21270"},"PeriodicalIF":8.4,"publicationDate":"2025-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145486514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：