Longitudinal–Lateral Decoupling Stability Control of Electric AHVs Based on BPNN Inverse System

ABSTRACT

In order to address the issues of instability arising from longitudinal‐lateral coupling during the motion of distributed electric drive AHVs (Articulated Heavy vehicles), a decoupling stability control method based on a BP neural network (BPNN) is proposed. The upper‐layer controller establishes an inverse system model of the vehicle's longitudinal and lateral dynamics based on inverse system decoupling theory and BPNN theory. It constructs a BPNN inverse system model with a network structure of 10–13–5, where the input is the vehicle state variables and the output is the required driving force and yawing moment. The lower‐layer controller employs a quadratic programming control algorithm to optimize torque distribution, with the objective of minimizing tire adhesion utilization. In addition, a Lyapunov‐based stability analysis was conducted for the closed‐loop error system composed of the BPNN inverse system and the PID controller, showing that the system possesses local practical stability within the operating conditions considered in this study and exhibits a certain degree of robustness against bounded disturbances. Finally, the effectiveness of the proposed control strategy was validated through co‐simulation based on the MATLAB/Simulink and TruckSim platform. The simulation results show that the vehicle dynamic responses both before and after decoupling are overall superior to those of the uncontrolled vehicle, while the decoupled vehicle exhibits better decoupling performance and stability‐control effectiveness. After decoupling, the peak absolute value of the articulation angle is reduced by 29.99% compared with that before decoupling, and the longitudinal velocity deviations of the tractor and trailer are further reduced by 55.41% and 43.69%, respectively. These results indicate that the proposed longitudinal–lateral decoupling controller can reduce the influence of lateral motion on longitudinal motion, achieve effective decoupling, and significantly improve the lateral stability of the distributed electric drive vehicle train.