Data-Driven Auto Tuning for Controller Gain of Path Following Control

Abstract

To realize mobile robot, the path following control method was proposed. Although this method is simple, the desired lane keeping can be realized. In this method, a controller gain must be tuned so that the tracking error is small. However, the feedback gain must be tuned by specific experiments nor iterations. In control theory field, a new auto-tuning method which is called as Virtual internal model tuning (VIMT) was proposed. VIMT can tune the controller by using one-shot experiment data. If this method can be applied to the vehicle, the feedback gain can be tuned. However, VIMT cannot be applied directly because the path following control is nonlinear control method. This study expands VIMT so as to tune the path following control. The proposed method can use VIMT by linearizing the path following control method. The validity of the proposed method is verified via numerical simulation.