Yuan Zou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Yuan Zou has a total of 130 co-authors. Here are the top co-authors by number of shared publications:
- Xudong Zhang
Beijing Collaborative and Innovative Center for Electric Vehicle and School of Mechanical Engineering, Beijing Institute of Technology, Beijing, China
32 shared publications - Guodong Du
School of Mechanical Engineering, Beijing Institute of Technology, Beijing, China
12 shared publications - Ningyuan Guo
School of Mechatronic Engineering and Automation, Guangdong Provincial Key Laboratory of Industrial Intelligent Inspection Technology, Foshan University, Foshan, China
10 shared publications - Tao Zhang
China North Vehicle Research Institute, Beijing 100072, China
9 shared publications Xiaosong Hu
Chongqing University
9 shared publicationsJunqiu Li
Beijing Institute of Technology
8 shared publications
Latest Publications
- Open Access
- 14 November 2026
- Published by Elsevier in Green Energy and Intelligent Transportation
Class-Aware Coupled Scene Interaction Network for Marginal Motion Forecasting in Interactive Driving Scenes
- 14 November 2026
- Published by Elsevier in Green Energy and Intelligent Transportation
Open AccessClass-Aware Coupled Scene Interaction Network for Marginal Motion Forecasting in Interactive Driving Scenes
- 9 July 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 332
ProgressiveDrive: Reconciling Fidelity and Consistency in World Models for Autonomous Driving
- 9 July 2026
- Published by Elsevier in Expert Systems with Applications
- Vol. 332
ProgressiveDrive: Reconciling Fidelity and Consistency in World Models for Autonomous Driving
- Top Cited Papers
- 17 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. PP (00189545) , 1-13
Integrated Cost-optimal Convex Optimization for Eco-Driving of Connected Hybrid Electric Vehicles on Sloped and Curved Roads
Sloped and curved roads are inevitable in daily driving, where slope and curvature act as critical external disturbances that directly influence energy-efficient driving performance. Meanwhile, an integrated eco-driving framework combining speed optimization and energy management can enhance energy-saving optimality. However, balancing computing burden and solution optimality for such frameworks under sloped and curved conditions remains challenging. To address this, this paper proposes an integrated convex optimization method for hybrid electric vehicles, aiming to improve driving efficiency, reduce energy consumption, and extend battery lifetime while ensuring high computational efficiency. First, a vehicle driving model incorporating road curvature, powertrain energy-flow model, and battery state-of-health update model, are established, forming an integrated eco-driving optimization problem targeting minimum total cost. Given the problem's high complexity, convex approximations and relaxations are applied to construct an integrated convex optimization problem in second-order cone programming form, balancing optimality and efficiency. The feasibility of the relaxation method is theoretically analyzed. Simulation tests on flat racetrack, viaduct, and hilly roads with slopes and curves show that the proposed method effectively reduces the total cost of energy savings and battery life extension, with negligible memory requirements and significantly higher computational efficiency, where computing time is less than 0.005% of that of dynamic programming.
- 17 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. PP (00189545) , 1-13
Top Cited PapersIntegrated Cost-optimal Convex Optimization for Eco-Driving of Connected Hybrid Electric Vehicles on Sloped and Curved Roads
Sloped and curved roads are inevitable in daily driving, where slope and curvature act as critical external disturbances that directly influence energy-efficient driving performance. Meanwhile, an integrated eco-driving framework combining speed optimization and energy management can enhance energy-saving optimality. However, balancing computing burden and solution optimality for such frameworks under sloped and curved conditions remains challenging. To address this, this paper proposes an integrated convex optimization method for hybrid electric vehicles, aiming to improve driving efficiency, reduce energy consumption, and extend battery lifetime while ensuring high computational efficiency. First, a vehicle driving model incorporating road curvature, powertrain energy-flow model, and battery state-of-health update model, are established, forming an integrated eco-driving optimization problem targeting minimum total cost. Given the problem's high complexity, convex approximations and relaxations are applied to construct an integrated convex optimization problem in second-order cone programming form, balancing optimality and efficiency. The feasibility of the relaxation method is theoretically analyzed. Simulation tests on flat racetrack, viaduct, and hilly roads with slopes and curves show that the proposed method effectively reduces the total cost of energy savings and battery life extension, with negligible memory requirements and significantly higher computational efficiency, where computing time is less than 0.005% of that of dynamic programming.
- 9 January 2026
- Published by Springer Nature in Automotive Innovation
- Vol. 9 (1) , 128-143
Task Offloading Algorithm Based on Preference Weight Adaptive Multi-objective Reinforcement Learning for Cloud-Assisted Mobile Edge Computing in Internet of Vehicles
- 9 January 2026
- Published by Springer Nature in Automotive Innovation
- Vol. 9 (1) , 128-143
Task Offloading Algorithm Based on Preference Weight Adaptive Multi-objective Reinforcement Learning for Cloud-Assisted Mobile Edge Computing in Internet of Vehicles
- Open Access
Task Offloading and Resource Allocation for ICVs in Vehicular Edge Computing Networks Based on Hybrid Hierarchical Deep Reinforcement Learning
Intelligent connected vehicles (ICVs) face challenges in handling intensive onboard computational tasks due to limited computing capacity. Vehicular edge computing networks (VECNs) offer a promising solution by enabling ICVs to offload tasks to mobile edge computing (MEC), alleviating computational load. As transportation systems are dynamic, vehicular tasks and MEC capacities vary over time, making efficient task offloading and resource allocation crucial. We explored a vehicle–road collaborative edge computing network and formulated the task offloading scheduling and resource allocation problem to minimize the sum of time and energy costs. To address the mixed nature of discrete and continuous decision variables and reduce computational complexity, we propose a hybrid hierarchical deep reinforcement learning (HHDRL) algorithm, structured in two layers. The upper layer of HHDRL enhances the double deep Q-network (DDQN) with a self-attention mechanism to improve feature correlation learning and generates discrete actions (communication decisions), while the lower layer employs deep deterministic policy gradient (DDPG) to produce continuous actions (power control, task offloading, and resource allocation decision). This hybrid design enables efficient decomposition of complex action spaces and improves adaptability in dynamic environments. Results from numerical simulations reveal that HHDRL achieves a significant reduction in total computational cost relative to current benchmark algorithms. Furthermore, the robustness of HHDRL to varying environmental conditions was confirmed by uniformly designing random numbers within a specified range for certain simulation parameters.
Open AccessTask Offloading and Resource Allocation for ICVs in Vehicular Edge Computing Networks Based on Hybrid Hierarchical Deep Reinforcement Learning
Intelligent connected vehicles (ICVs) face challenges in handling intensive onboard computational tasks due to limited computing capacity. Vehicular edge computing networks (VECNs) offer a promising solution by enabling ICVs to offload tasks to mobile edge computing (MEC), alleviating computational load. As transportation systems are dynamic, vehicular tasks and MEC capacities vary over time, making efficient task offloading and resource allocation crucial. We explored a vehicle–road collaborative edge computing network and formulated the task offloading scheduling and resource allocation problem to minimize the sum of time and energy costs. To address the mixed nature of discrete and continuous decision variables and reduce computational complexity, we propose a hybrid hierarchical deep reinforcement learning (HHDRL) algorithm, structured in two layers. The upper layer of HHDRL enhances the double deep Q-network (DDQN) with a self-attention mechanism to improve feature correlation learning and generates discrete actions (communication decisions), while the lower layer employs deep deterministic policy gradient (DDPG) to produce continuous actions (power control, task offloading, and resource allocation decision). This hybrid design enables efficient decomposition of complex action spaces and improves adaptability in dynamic environments. Results from numerical simulations reveal that HHDRL achieves a significant reduction in total computational cost relative to current benchmark algorithms. Furthermore, the robustness of HHDRL to varying environmental conditions was confirmed by uniformly designing random numbers within a specified range for certain simulation parameters.
- 30 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (24) , 53288-53307
Joint Task Offloading and Resource Allocation for Vehicle Platoon: A Nested Algorithm Based on Stackelberg Game
With the increasing computational demands of intelligent connected vehicles (ICVs), traditional roadside mobile edge computing (MEC) faces resource and coverage limitations. As a mobile computing entity, a vehicle platoon enables coordinated resource scheduling and low-latency communication, serving as an effective supplement to edge computing. However, limited onboard resources make vehicle platoons insufficient for handling heavy task loads. This study focuses on the joint task offloading and resource allocation between vehicle platoons and MEC servers. An MEC-assisted vehicle platoon computing network (MVPCN) is constructed, where both platoon vehicles (PVs) and MEC servers serve as computing nodes. We formulate a joint optimization problem of task offloading and resource allocation in the vehicle platoon, aiming to minimize the long-term weighted cost of time and energy consumption. Given the complexity of the problem and the limitations of conventional solution methods, we design a discrete two-stage Stackelberg game (DTSG) to decompose the problem into two subproblems: leader [platoon leader (PL)] level and follower (computing nodes) level and prove the existence of a Stackelberg equilibrium (SE). Based on the game model, we propose a nested platoon task offloading and resource-allocation (NPTORA) algorithm. The leader employs the multiagent proximal policy optimization (MAPPO) to generate offloading decisions for the platoon, while the followers allocate computational resources using a dynamic priority-based hybrid optimization (DPHO) algorithm based on the received offloading decisions. Simulation results show that NPTORA outperforms baseline methods in task success, cost efficiency, and adaptability, demonstrating strong performance and practical potential.
- 30 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (24) , 53288-53307
Joint Task Offloading and Resource Allocation for Vehicle Platoon: A Nested Algorithm Based on Stackelberg Game
With the increasing computational demands of intelligent connected vehicles (ICVs), traditional roadside mobile edge computing (MEC) faces resource and coverage limitations. As a mobile computing entity, a vehicle platoon enables coordinated resource scheduling and low-latency communication, serving as an effective supplement to edge computing. However, limited onboard resources make vehicle platoons insufficient for handling heavy task loads. This study focuses on the joint task offloading and resource allocation between vehicle platoons and MEC servers. An MEC-assisted vehicle platoon computing network (MVPCN) is constructed, where both platoon vehicles (PVs) and MEC servers serve as computing nodes. We formulate a joint optimization problem of task offloading and resource allocation in the vehicle platoon, aiming to minimize the long-term weighted cost of time and energy consumption. Given the complexity of the problem and the limitations of conventional solution methods, we design a discrete two-stage Stackelberg game (DTSG) to decompose the problem into two subproblems: leader [platoon leader (PL)] level and follower (computing nodes) level and prove the existence of a Stackelberg equilibrium (SE). Based on the game model, we propose a nested platoon task offloading and resource-allocation (NPTORA) algorithm. The leader employs the multiagent proximal policy optimization (MAPPO) to generate offloading decisions for the platoon, while the followers allocate computational resources using a dynamic priority-based hybrid optimization (DPHO) algorithm based on the received offloading decisions. Simulation results show that NPTORA outperforms baseline methods in task success, cost efficiency, and adaptability, demonstrating strong performance and practical potential.
- 11 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (22) , 48945-48961
LUOT: LiDAR–UWB Object Tracking With Zero ID Switches and Centimeter-Level Precision
Scenarios such as vehicle platooning urgently require high-precision object tracking algorithms with strong identity (ID) consistency. However, current state-of-the-art (SOTA) 3-D tracking methods often suffer from ID switches. To address this limitation, this article proposes a light detection and ranging (LiDAR)–ultrawideband (UWB) object tracking (LUOT) algorithm designed to achieve centimeter-level ($1\sigma $ ) 3-D object tracking with zero ID switches. The method fuses high-precision LiDAR-based bounding boxes (BBoxes) with the ID consistency of UWB. Considering the high precision of 3-D BBoxes at positions densely populated by point clouds, LUOT replaces the conventional association center point with coordinates transformed to the vehicle coordinate frame through the vehicle’s rear-end frame, where point clouds are denser, thereby minimizing errors arising from center point detection inaccuracies. To mitigate the destructive impact of occasional UWB outliers, this study pioneers the introduction of a Mahalanobis distance-based chi-square test, integrating multiple UWB ranges to comprehensively detect and eliminate outliers. Finally, real-world vehicle experiments demonstrate that LUOT achieves zero ID switches with a tracking accuracy of 0.0868 m ($1\sigma $ ), fully satisfying centimeter-level precision requirements for vehicle autonomous driving. These results establish LUOT as an SOTA solution for vehicle-to-vehicle global object tracking and fill an important gap in LiDAR–UWB fusion methods. The source code is publicly available at https://github.com/ly3106/LUOT- 11 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 12 (22) , 48945-48961
LUOT: LiDAR–UWB Object Tracking With Zero ID Switches and Centimeter-Level Precision
Scenarios such as vehicle platooning urgently require high-precision object tracking algorithms with strong identity (ID) consistency. However, current state-of-the-art (SOTA) 3-D tracking methods often suffer from ID switches. To address this limitation, this article proposes a light detection and ranging (LiDAR)–ultrawideband (UWB) object tracking (LUOT) algorithm designed to achieve centimeter-level ($1\sigma $ ) 3-D object tracking with zero ID switches. The method fuses high-precision LiDAR-based bounding boxes (BBoxes) with the ID consistency of UWB. Considering the high precision of 3-D BBoxes at positions densely populated by point clouds, LUOT replaces the conventional association center point with coordinates transformed to the vehicle coordinate frame through the vehicle’s rear-end frame, where point clouds are denser, thereby minimizing errors arising from center point detection inaccuracies. To mitigate the destructive impact of occasional UWB outliers, this study pioneers the introduction of a Mahalanobis distance-based chi-square test, integrating multiple UWB ranges to comprehensively detect and eliminate outliers. Finally, real-world vehicle experiments demonstrate that LUOT achieves zero ID switches with a tracking accuracy of 0.0868 m ($1\sigma $ ), fully satisfying centimeter-level precision requirements for vehicle autonomous driving. These results establish LUOT as an SOTA solution for vehicle-to-vehicle global object tracking and fill an important gap in LiDAR–UWB fusion methods. The source code is publicly available at https://github.com/ly3106/LUOT