Simulation and experiments on preview tracking controller for intelligent vehicles based on hierarchical and refined estimation of multi-heterogeneous disturbances

The path tracking control system of intelligent vehicles is affected by multiple heterogeneous disturbances, including modeling inaccuracies, network delays, parameter perturbations, and external environmental factors. Addressing these influences is critical to achieving precise tracking. This paper proposed a new robust preview control framework with hierarchical and refined disturbance rejection, which integrates linear quadratic regulator, hierarchical and refined estimation of disturbances, and adaptive preview model into a control system with efficient disturbance rejection. Firstly, an adaptive preview model with advanced predictability is designed to obtain the states of the future target “road,” effectively handling the unavoidable network delays. On this basis, a preview error dynamic model is established considering multi-heterogeneous disturbances. According to the heterogeneity of the disturbances, the multiple disturbances can be refined into two parts: the known external time-varying disturbances caused by road curvature and the unknown disturbances caused by other factors. These are estimated using a preview forward feedback controller (PFFC) and a linear extended state observer (LESO), respectively. After that, the path tracking problem is designed as an optimal controller based on a linear quadratic regulator (LQR) to obtain the steering wheel angle with safety constraints to track the desired path. The asymptotic stability of the closed-loop system is analyzed through the Lyapunov theorem. Finally, the proposed controller is verified and compared by the Carsim-Simulink simulation platform preliminarily, and real-vehicle experiments are implemented on an intelligent vehicle platform. Simulation and real-vehicle experimental results show excellent control performance.