Performance comparison of controllers’ design on trajectory tracking of public autonomous vehicle based on genetic algorithm

This research proposes novel methods of designing Sliding Mode (SMC), Fuzzy Logic (FLC) controllers to control the vehicle for tracking various trajectories. The first contribution is the development of the SMC design method applying the Decoupled Algorithm and the Hurwitz condition to minimize the number of control variables and accurately determine relationships between sliding surface parameters. The second contribution is an FLC design method with completely new fuzzy rules that ensure the lateral tracking by only two input variables (lateral ey and heading eϕ errors). Additionally, the study applies ANFIS to develop a single controller, PAC, capable of operating effectively under various scenarios without complex parameter updates, using FLC’s data. The scenarios used in the study include closed-loop and short open-loop routes simulating vehicle’s typical routes designed with Matlab Driving-Scenario Designer. Compared with the FLC–Trimf–Gauss controller, the SMC achieved significant performance, reducing the RMS ey and RMS eϕ by 45.5% and 30.7%, respectively; the FLC–Trimf and PAC controllers showed nearly the same performance, with reductions in RMS ey and RMS eϕ of 27.3% and 16.7%, respectively.