Fixed-time self-structuring neural network fault-tolerant tracking control of underactuated surface vessels with state constraints

Abstract

In this paper, a fixed-time fault-tolerant control strategy is proposed for trajectory tracking of underactuated surface vessels (USVs) with input saturation. A predefined region boundary function is proposed, which has all predefined fixed-time convergence properties and provides predefined specifications for tracking errors. To achieve the predetermined tracking performance, a fixed-time adaptive neural network state constrained fault-tolerant controller is proposed based on dynamic surface control technology, a back-stepping program, and an asymmetric time-varying tangent barrier Lyapunov function (ATVTBLF). The unknown external environment disturbances and complex continuous unknown nonlinear terms are approximated by a self-structuring neural network (SSNN), which can improve the robustness of the control system to unknown parameters and guarantee the tracking performance of the control system. The number of neurons of the SSNN can be adjusted online according to the approximation efficiency, which can reduce the amount of calculation of the control system. Based on Lyapunov stability theory, it is proven that the closed-loop system is bounded and stable in a fixed time, and the tracking error is always kept within the constraint range. The simulation results show that the underactuated surface vessel control system has good tracking performance.