Gaussian-process-based shared-ICR adaptive tube MPC for trajectory tracking of all-axle steer-by-wire heavy vehicles

Abstract

All-axle steer-by-wire (ASBW) systems enable independent steering control of all axles and can improve the trajectory tracking performance of heavy vehicles. However, during heavy-load transportation, frequent load variations intensify parameter perturbations, such as tire cornering stiffness variations, under independent axle steering. In addition, when the vehicle approaches critical instability, the tire-force nonlinearities of different axles become more pronounced, making optimal coordination of all axle steering angles difficult and severely limiting trajectory tracking performance. Therefore, this paper proposes a shared instantaneous center of rotation (shared-ICR) adaptive trajectory tracking control method for ASBW heavy vehicles. On the one hand, a shared-ICR steering scheme is developed by introducing a mathematical variable-transformation strategy, which converts the multi-axle steering coordination problem into a low-dimensional ICR-variable optimization problem, thereby reducing control complexity and providing a basis for optimal coordination of all axle steering angles. On the other hand, an adaptive Tube model predictive control (Tube MPC) trajectory tracking controller is designed, where the effects of parameter perturbations and nonlinearities are collectively modeled as ICR-related disturbances. A Gaussian process (GP) method is employed to adaptively characterize the boundary of ICR-related disturbances and construct dynamic Tube cross-sections. Moreover, an ICR stability-margin scaling factor is designed to adjust the Tube cross-sections online, realizing adaptive regulation with respect to vehicle stability variations and improving trajectory tracking performance under disturbances and near-critical instability conditions. The results demonstrate that the proposed method improves the trajectory tracking accuracy and stability of ASBW heavy vehicles.