Path Tracking Control of Hierarchical and Refined Active Disturbance Rejection for Intelligent Vehicles Under Heterogeneous Disturbances

Intelligent vehicle path tracking systems are a class of nonlinear systems with many uncertainties such as modeling, parameters perturbation, speed fluctuation, external environmental disturbances and data packet dropout. The elimination of heterogeneous disturbances brought by the above factors is the key to achieve accurate and stable tracking. This article accordingly proposes a hierarchical and refined active disturbance rejection control (ADRC) method considering heterogeneous disturbances and their response characteristics in the path tracking system. Firstly, the disturbances-incorporated model for path tracking is established based on the lateral and yaw dynamic responses of vehicle, and a linear extended state observer (LESO) is designed for estimating the system disturbance, and a closed-loop feedback control is realized by constructing the ADRC framework. On this basis, the problem of LESO and the system steady-state disturbance are further analyzed, and the robust compensation term based on radial basis function neural networks (RBFNN) is designed and acted on the system feedforward layer. Ultimately, the parallel two-layer composite ADRC-RBFNN framework with refined anti-disturbance and consistent boundedness is obtained. Simulation and hardware-in-loop (HiL) results show the superiority of the proposed controller in disturbance rejection, which can effectively improve the stability margin and has good tracking accuracy, robustness and real-time performance.