Intelligent Vehicle Robust Trajectory Tracking Algorithm Based on Fuzzy Adaptive Dynamic Model Predictive Control

Model predictive control (MPC) has been widely applied in the field of autonomous driving; however, it still suffers from poor control accuracy in extreme conditions. To improve the smoothness and robustness of intelligent vehicle trajectory tracking, this study tries to investigate an improved MPC trajectory tracking control for intelligent vehicle, fuzzy adaptive dynamic model predictive control (FADMPC), with prediction time-domain dynamic optimization and objective function weight adaptive regulation. First, in the improved MPC framework, a vehicle dynamics model with three degrees of freedom is established as the theoretical vehicle model of MPC controller, and the recursive least square method (RLS) with forgetting factor is used to timely estimate tire cornering stiffness considering nonlinear error of the control-oriented vehicle model. Second, the relationship between the prediction time domain of the model and factors such as vehicle speed and road adhesion coefficient is analyzed through simulation. Then, a dynamic time-domain regulation scheme is determined by using the whale optimization algorithm (WOA). Third, a fuzzy controller is used to dynamically adjust the weight coefficient of the lateral displacement tracking error in the objective function. Last, a series of experiments is performed on a hardware-in-the-loop platform to validate the comparison effect of different trajectory tracking controllers on jointed and bisected adhesion condition roads. Results indicate that the proposed FADMPC can effectively improve trajectory tracking accuracy and vehicle stability compared with LQR and even ADMPC controller.