Multiple sliding mode adaptive fuzzy controller for nonlinear marine autopilot systems

When a ship course controller must deal with unknown parameters,the maneuvers of the ship must be described using an unmatched and uncertain nonlinear mathematical model with unknown virtual control coefficients and unknown parameters. To deal with this situation,an adaptive fuzzy control algorithm with multiple sliding modes was proposed. It combined multiple sliding mode design technology with the approximation capability of a fuzzy logic system. In order to solve the problem of parameter drift and the controller's singularity problem,a non-continuous projection algorithm and an integral-type Lyapunov function were employed in the proposed algorithm. By means of the Lyapunov function,it was theoretically proven that the proposed controller makes all signals in the resulting closed-loop adaptive system uniformly bounded,with tracking error converging to zero. The simulation results showed that,compared to conventional proportional-integral-derivative (PID) controllers,the controller discussed in this paper provided better adaptability and tracking capabilities.