Improved Fractional-Order Integral Sliding Mode Control for AUV Based on RBF Neural network

Abstract

Aiming at the high control accuracy and stability requirements of the autonomous underwater robot (AUV) in underwater docking process under complex sea conditions, a robust adaptive control algorithm combining RBF neural network and improved fractional integral sliding mode control is proposed. To suppress the influence of parameter perturbation, the RBF neural network is used to approximate the uncertain interference and the uncertainty of the AUV model. At the same time, the fractional order integral is used to improve the large overshoot caused by the ordinary integral sliding mode control. In order to make the system state quickly converge to the fractional sliding mode surface without chattering, a new double power approach law with fast convergence is proposed. By Lyapunov theorem, A stability analysis proves that the system can quickly converge to a neighborhood of the origin. At last, simulation results illustrate the effectiveness and robustness of the proposed controller.