Application of Robust Optimal Sliding Mode Control for Underactuated AUV System with Uncertainties

To realize the robust control of underactuated Autonomous Underwater Vehicle (AUV) with uncertainties of un-modeled dynamics, parameter perturbations and random disturbances, a robust optimal sliding mode control approach is proposed based on Linear quadratic optimal regulator (LQR) and sliding mode control. Firstly, the AUV model in vertical plane is given. Secondly, based on the State Dependent Riccati Equation (SDRE), the optimal controller is designed for the AUV nominal system without uncertainties, which makes the quadratic performance index optimal. Then, considering the uncertainty of the system, the robust optimal integral sliding mode control law is designed based on SDRE nominal controller, which makes the AUV system satisfy the requirement of the performance index, and it is robust to the parameter perturbations and random disturbances. Finally, the validity and robustness of the presented control approach is demonstrated by a dynamic model of the Remote Environmental Unit (REMUS) AUV.

1 Introduction

Autonomous Underwater Vehicle (AUV) has become one of the research hotspots in the control field because of its wide application background in the fields of ocean development, seabed exploration and military [1]-[14]. At present, there are many control methods for AUV, such as PID method[1], neural network control method[2], adaptive controller[3]-[4], optimal control method[5]-[8], sliding mode control method[9]-[12], etc. AUV system has strong nonlinearity, hydrodynamic coefficient uncertainty and external disturbance. These uncertainties require strong robustness of the controller. Sliding mode control is more robust and invariant to any perturbation and external disturbance satisfying matching conditions, and has fast response ability. Therefore, it is often used in AUV control. In reference [1], a self-adaptive fuzzy PID controller is designed for AUV in heading and depth attitude. In reference [2]-[3], a DRFNN-adaptive control method is proposed to solve the problems of depth tracking of AUV; In reference [4], the formation control problem is studies, and an adaptive controller is designed for AUV with model uncertainties. In reference [5]-[8], the State-dependent Riccati Equation approach is applied for solving the control problem for AUV with control constraints. an adaptive fuzzy sliding mode controller is designed to deal with the depth control of remotely operated underwater vehicles in reference [9],[10]. In reference [11],[12], the sliding mode control approach is prosed for the trajectory tracking of underactuated AUV. An approximate optimal tracking control approach is presented for near-surface AUVs with wave disturbances in [13]. Among the existing control methods, some specific problems for AUV have been solved.