ADP-Based Optimal Stabilization for Ship Autopilot under Actuator Faults

Abstract

The safe and energy-saving operation of ship autopilots has attracted extensive research attention in marine engineering. This paper addresses the stabilization control of ship autopilots via adaptive dynamic programming (ADP) when velocity measurements are unavailable and actuator faults occur. A robust ADP scheme is developed to achieve fault-tolerant stabilization and performance optimization merely using heading angle signals. First, an adaptive uncertainty observer is constructed to estimate the unknown heading angular rate and actuator faults. Then, the ADP framework is employed to approximate the system’s unknown dynamics. By means of online learning, the optimization problem for the error dynamics is resolved, which minimizes the cost function while maintaining satisfactory steady-state control performance. The Lyapunov stability analysis rigorously proves that all signals in the closed-loop system are uniformly ultimately bounded. Numerical simulations further validate the superior performance of the presented control scheme.