Path tracking of an autonomous vehicle by means of an indirect adaptive neural controller

Abstract

Due to nonlinearity of the dynamic models and the advantages of neural networks, we design an indirect adaptive neural controller for tracking the path of an autonomous vehicle. Our proposed algorithm, which includes a neural identifier, controls the lateral movement of the autonomous car. The updating laws for the identifying neural network and the controlling neural network are obtained by means of the gradient descent method. The identifying neural network is employed to estimate the Jacobian of the vehicle's lateral movement, which is then used for updating the parameters of the adaptive neural controller online. In this paper, both the radial basis function (RBF) and the multilayer perceptron (MLP) types of indirect adaptive neural controllers have been deigned. The simulation results corroborate the satisfactory performance of our proposed method in controlling the lateral motion of the examined autonomous vehicle. The results of RBF and MLP indirect adaptive neural controllers have also been compared in this work.