Adaptivity-Enhanced Path Tracking System for Autonomous Vehicles at High Speeds

The path tracking is a basic function of the autonomous vehicle. Generally, it is difficult to realize the path tracking with high accuracy at high speeds. To solve this problem, an adaptivity-enhanced path tracking system (AEPTS) is investigated in this paper. The adaptive performance of this AEPTS is realized via three aspects. The first is to achieve preview distance adaptivity by a dynamic multi-point preview model (DMPPM), which provides the desired course deviation according to the vehicle speed and the reference path. The second is to improve the adaptivity of an adaptive fuzzy steering controller (AFSC) with an adaptive steering gainmodel which is built via the path curvature and vehicle speed. The third is that an optimal parameter surface is devised to enhance the adaptivity of the AEPTS at high speeds. Such a surface can provide the optimal coefficients for the AEPTS, and it is implemented using the particle swarm optimization (PSO) and a curve fitting method. At last, simulation experiments are given to verify the adaptivity of the proposed method.