YJYuan Ji

Yuan Ji

*This is an algorithmically generated author record based on publicly available information.

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


Yuan Ji has a total of 24 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Interactive Motion Planning Using Genetic-Enhanced Reinforcement Learning With Backup Strategy

    by Xiaohui Hou, Minggang Gan, Wei Wu, Yuan Ji, Shiyue Zhao, Jie Chen

    Navigating dense traffic environments, particularly during unprotected left-turn maneuvers, poses a critical challenge for autonomous vehicles due to the uncertainty of traffic participants, the trade-off between safety and efficiency, and dynamic interaction between vehicles. Existing methods usually fail to meet the constraints in changing environments or result in overly cautious behavior. This paper proposes a Genetic-Enhanced Reinforcement Learning with Backup Strategy (GERL-BS) to address these challenges by effectively navigating unprotected left turns in dense traffic. By leveraging the evolutionary capabilities of a risk-concerned Genetic Algorithm (GA), GERL-BS enhances the exploration of safety critical states by distilling essential interaction knowledge and control strategies, overcoming the limitations of Reinforcement Learning (RL) in such high-risk scenarios. Additionally, the Safety-Enhanced Contingency Backup (SECB) Module introduces a reward augmentation mechanism to ensure safety in complex and uncertain environments. To address the diversity and variability in vehicle interactions, we introduced the Heterogeneous Intelligent Driver Model (H-IDM), designed to simulate the heterogeneous behaviors of neighboring vehicles. Comprehensive evaluations in dense traffic scenarios with varying vehicle configurations demonstrate the proposed GERL BS controller's superior safety, adaptability, and effectiveness.

  • Equipping With Cognition: Interactive Motion Planning Using Metacognitive-Attribution Inspired Reinforcement Learning for Autonomous Vehicles

    by Xiaohui Hou, Minggang Gan, Wei Wu, Yuan Ji, Shiyue Zhao, Jie Chen

    This study introduces the Metacognitive-Attribution Inspired Reinforcement Learning (MAIRL) approach, designed to address unprotected interactive left turns at intersections—one of the most challenging tasks in autonomous driving. By integrating the Metacognitive Theory and Attribution Theory from the psychology field with reinforcement learning, this study enriches the learning mechanisms of autonomous vehicles with human cognitive processes. Specifically, it applies Metacognitive Theory’s three core elements—Metacognitive Knowledge, Metacognitive Monitoring, and Metacognitive Reflection—to enhance the control framework’s capabilities in skill differentiation, real-time assessment, and adaptive learning for interactive motion planning. Furthermore, inspired by Attribution Theory, it decomposes the reward system in RL algorithms into three components: 1) skill improvement, 2) existing ability, and 3) environmental stochasticity. This framework emulates human learning and behavior adjustment, incorporating a deeper cognitive emulation into reinforcement algorithms to foster a unified cognitive structure and control strategy. Contrastive tests conducted in various intersection scenarios with differing traffic densities demonstrated the superior performance of the proposed controller, which outperformed baseline algorithms in success rates and had lower collision and timeout incidents. This interdisciplinary approach not only enhances the understanding and applicability of RL algorithms but also represents a meaningful step towards modeling advanced human cognitive processes in the field of autonomous driving.


  • Risk-Conscious Mutations in Jump-Start Reinforcement Learning for Autonomous Racing Policy

    by Xiaohui Hou, Minggang Gan, Wei Wu, Shiyue Zhao, Yuan Ji, Jie Chen

    This study focuses on trajectory planning and motion control policies in autonomous racing, which necessitates pushing the capacity boundaries of racing vehicles to achieve maximum speeds and minimal lap times. We propose an innovative planning control framework that integrates risk-conscious mutations in jump-start reinforcement learning (RCM-JSRL) and nonlinear model predictive control (NMPC). The RCM-JSRL algorithm incorporates jump-start curriculum learning and the risk-conscious genetic algorithm into reinforcement learning, leveraging prior expert knowledge and a curiosity-driven exploration mechanism to enhance training efficiency while avoiding excessively conservative policy generation in high-complexity and high-risk scenarios. NMPC generates locally optimal control commands that adhere to vehicle dynamics constraints while following the designated trajectory. Following training on track maps with varying difficulty levels, the proposed controller successfully executes a superior policy compared to the guide policy, providing evidence of its effectiveness and scalability. It is our belief that this technology can be applied in everyday driving scenarios, improving efficiency under special conditions, ensuring stability in critical situations, and broadening the scope of autonomous driving applications.


  • Predictive Wheel Cylinder Pressure Control for Automotive Hydraulic Brake-by-Wire Systems

    by Weilong Liu, Junzhi Zhang, Ruihai Ma, Yuan Ji, Xiaoxia He, Chengkun He

    High-speed on/off valve (HSV), characterized by its low cost, fast response, and strong robustness, has been extensively used in automotive hydraulic brake-by-wire systems. Studies on wheel cylinder pressure (WCP) control methods based on HSVs have been ongoing for a long time. To alleviate the calibration workload and avoid the usage of the pulsewidth modulation technique, this article proposes an intuitive and easy-to-implement predictive pressure control (PPC) method by directly taking advantage of the on/off nature of the HSV. First, considering the HSV as an ideal hydraulic switch with two states of on and off , the finite control combinations of the inlet and outlet valves are defined. Subsequently, the corresponding finite flow rate set WCP model is proposed. Second, the principle and design of the proposed PPC method are presented. Based on pressure measurements and disturbance observation, at every sampling instant, the WCPs are predicted by applying all control combinations to the finite flow rate set WCP model. These predictions are evaluated with a cost function that minimizes the pressure tracking error. Then, the control combination with minimum cost is selected and applied. Finally, a hardware-in-the-loop test bench is set up to implement and validate the proposed PPC method. The experimental results demonstrate that compared to a classic WCP control method based on the pulsewidth modulation technique, the proposed PPC method improves precision by at least 10% and 40% in the steady-state and dynamic pressure tracking experiments for normal braking demands, respectively.


  • Cross-Observability Optimistic-Pessimistic Safe Reinforcement Learning for Interactive Motion Planning With Visual Occlusion

    by Xiaohui Hou, Minggang Gan, Wei Wu, Yuan Ji, Shiyue Zhao, Jie Chen

    This study focuses on the motion planning and risk evaluation of unprotected left turns at occluded intersections for autonomous vehicles. In this paper, we present an interactive motion planning controller that combines Cross-Observability Optimistic-Pessimistic Safe Reinforcement Learning (COOP-SRL) and Nonlinear Model Predictive Control (NMPC), with consideration of the uncertain potential risk of occluded zone, the trade-off between safety and efficiency, and the dynamic interaction between vehicles. The proposed COOP-SRL algorithm integrates fully and partially observable policies through cross-observability soft imitation learning to leverage the expert guidance and improve learning efficiency. Moreover, the optimistic exploration policy and pessimism safe constraint are adopted to provide an adaptive safe strategy without hindering the exploration during learning process. Finally, the evaluations of the proposed controller were conducted in occluded intersection scenarios with various traffic density level, which indicate that the proposed method outperforms both the optimization-based and learning-based baselines in qualitative and quantitative indexes.


  • Safe Replanning and Motion Control for Automated Emergency Braking Under Braking System Failures

    by Jinheng Han, Junzhi Zhang, Chen Lv, Chengkun He, Henglai Wei, Shiyue Zhao, Zhongshi Zhang, Yuan Ji

    The automated emergency braking system is designed to prevent rear-end collisions. However, unexpected failures within this system during braking could potentially lead to accidents. To address this concern, we propose an integrated safe replanning and motion control framework to mitigate collision risks in the event of braking failures. Unlike previous fault-tolerant control works, the steering system is intervened to eliminate failure risks instead of re-configuring the braking torque; therefore, the longitudinal driving risks can be lessened from the lateral perspective. To achieve this approach, firstly, a risk assessment function with a failure factor is constructed to reconsider the driving target. Then, the driving target position and surrounding obstacle information are incorporated into one novel integral heuristic barrier Lyapunov function to reformulate the motion replanning and tracking problems as one yaw angle tracking problem. After, a novel barrier function-based sliding mode controller is introduced to ensure the yaw angle tracking performance within constraints. In this paper, decision-making, motion planning, and tracking tasks are incorporated into one analytical solution framework uniformly.Additionally, the prescribed time stability of our strategy ensures an upper bound setting time, indicating stability before the ego vehicle stop completly. Finally, numerous simulation results are carried out to confirm the effectiveness and superiority of our strategy in circumventing the troubles of braking system failures.


  • Optimal Path Tracking Control Based on Online Modeling for Autonomous Vehicle With Completely Unknown Parameters

    by Yuan Ji, Junzhi Zhang, Chen Lv, Chengkun He, Hao Chen, Jinheng Han, Xiaohui Hou

    Reliable path tracking control (PTC) method is essential for autonomous driving. However, existing PTC methods count on prior vehicle parameters to achieve good performance. This paper presents an optimal PTC method without requiring any prior vehicle parameters based on online modeling with strict parameter convergence ability. First, we build a virtual optimal control problem using adaptive dynamic programming (ADP) scheme to guide the data collection and solve two characteristic matrices containing parameter information. Then, the model construction method is derived using the solved matrices and the optimal PTC method is constructed using the constructed model. Finally, a fault-tolerant control scheme is further designed using the constructed model and the online modeling ability of the proposed method. The effectiveness of the proposed method is validated through co-simulation between Matlab/Simulink and high-fidelity vehicle dynamic simulation software CarSim $^\circledR$ under both fault-free and fault-tolerant situations.


  • Fault-Tolerant Vehicle Stability Control Based on Active Steering and Direct Yaw Moment With Finite-Time Constraint Performance Recovery

    by Yuan Ji, Junzhi Zhang, Chen Lv, Chengkun He, Xiaohui Hou, Jinheng Han

    Vehicle lateral stability is of great importance for maintaining vehicle safety. Active front-wheel steering (AFS) is the typical actuator to realize lateral control. However, if a fault occurs on AFS, direct yaw rate control (DYC) can also be used to control lateral stability through differential braking, guaranteeing actuation redundancy. The paper proposes a practical active fault tolerant method for controlling the AFS to DYC process in case of a complete failure of AFS, whereby DYC is expected to take over the control after a certain duration. The proposed method has a distinctive feature in that it can keep the yaw rate tracking error within a predetermined boundary before the failure occurrence and recover the error back to the preset boundary within a desired time period after the DYC takes over. Specifically, the proposed method can guarantee a strict constraint of yaw rate tracking error before the fault occurs on AFS. Furthermore, the method can recover the error back to the desired error boundary in a finite time once the DYC starts to operate. The key points of the method are as follows: First, a tangent barrier Lyapunov function (TBLF) is used with a backstepping mechanism to strictly constrain the yaw rate tracking error in a desired error boundary. Second, the delayed switch from AFS to DYC, which may cause a violation of the preset error boundary, is considered. And a set of functions named reconfigurable transform functions (RTF) are carefully designed, using which the tracking error can recover to the preset boundary within the desired time. The effectiveness of the proposed method is validated and compared with two comparative methods through co-simulation between the High-fidelity vehicle simulation software CarSim and Matlab/Simulink.