Intelligent commercial vehicle trajectory planning method with adaptive vehicle stability boundary

During the operation of commercial vehicles, the adhesion conditions vary frequently and the load changes dynamically, resulting in time-varying nonlinear characteristics of vehicle stability Boundary. Traditional trajectory planning methods exhibit weak and fixed dynamic boundary constraints, making it challenging to ensure the stability and comfort of intelligent commercial vehicles in scenarios with time-varying vehicle stability Boundary. To address this issue, this paper proposes a Model Prediction Trajectory Planning Method for Adaptive Vehicle Stability Boundary (MPTP-AVSB). On the one hand, by estimating the mass of commercial vehicles and road adhesion characteristics, a Lyapunov stability potential energy function for commercial vehicles is constructed to achieve precise quantitative characterization of the dynamic stability boundary of commercial vehicles, thereby providing dynamic and accurate stability boundary constraints for their trajectory planning. On the other hand, a time-varying model prediction trajectory planner adaptive to vehicle stability Boundary is proposed, which dynamically adjusts the planner’s state-space matrix and boundary constraints upon changes in the vehicle stability boundary. Additionally, a dynamic weight adjustment method based on deep reinforcement learning is designed to balance objectives such as trajectory stability, safety, and comfort, ultimately generating a highly stable and smooth trajectory. The results demonstrate that, compared with existing methods, the proposed MPTP-AVSB method generates trajectories with superior comfort and stability in scenarios with time-varying vehicle stability boundary.