An LS-SVM control method for path following of autonomous ground vehicles

In this paper, we investigate the Least Squares-Support Vector Machine (LS-SVM) control for path following problem of autonomous ground vehicles (AGV). Firstly, the steering angle input and the lateral offset output of the path following system model which is the two degree of freedom (DOF) vehicle model with integrated vision model are collected to form the training data of the LS-SVM model. Then, the LS-SVM model for the path following of vehicle is trained from the collected data and the controller is solved from the LS-SVM model. The effectiveness of the proposed control approach is validated by the performance of the closed-loop path following system in Matlab/Simulink.