Path tracking control of high-speed intelligent vehicles considering model mismatch

The precision of path tracking in high-speed intelligent vehicles is significantly influenced by model mismatch arising from factors like parameter uncertainty, model simplification, external disturbances, and other sources. In this paper, we propose a novel robust control strategy that integrates the compensation function observer (CFO) with the model predictive control (MPC) method, utilizing an optimized vehicle dynamics model (opt-model) to address this challenge, called OCMPC. Initially, we establish the opt-model to design predictive model by leveraging suspension kinematics and compliance (K&C) data collected from a miniature pure electric vehicle. Remarkably, the opt-model exhibits improved accuracy compared to the conventional vehicle dynamics model (con-model) while preserving the same degrees of freedom (DOF). Next, we incorporate CFO into the path tracking process of high-speed intelligent vehicles, enabling dynamic real-time observation of the model mismatch between the prediction model and the actual vehicle. CFO can capture the dynamics of the vehicle, including nonlinearities and uncertainties, without placing a heavy computing burden on the controller. This observed mismatch is subsequently employed for feed-forward compensation, facilitating the attainment of optimal control values. Ultimately, we validate the effectiveness of our proposed method in enhancing path tracking accuracy for high-speed intelligent vehicles through co-simulation using Simulink and Carsim.