Fixed-time Trajectory Tracking Control for Underactuated Surface Vehicle Based on Sliding Mode and Neural Network

Abstract

In recent years, underactuated surface vehicle’s trajectory tracking has become an academic hotspot. Since the independent control input of the underactuated surface craft is less than the freedom of motion, the controller design of the underactuated surface craft is relatively difficult. In addition, the nonlinearity of the surface vehicle model and the unknown disturbance of the ocean environment also make the high-precision trajectory tracking control design more difficult. In this paper, firstly the trajectory tracking error is redefined benefiting from the output redefinition-based dynamic transformation (ORDT) to construct a relative order system and simplify the design process of control law. Secondly, a fixed-time sliding mode control (FTSMC) is designed, in which both surge and yaw control are designed in one vector to achieve a fixed-time bounded trajectory tracking error. Thirdly, a radial basis function-based neural network (RBFNN) is designed to estimate complex fluid damping, unknown marine environmental disturbances, and unmodeled dynamics, and the complexity of controller design is reduced by means of the minimum learned parameter method (MLP). At last, the validity of control method design is validated by numerical simulation.