ADAPTIVE NEURAL TRACKING CONTROL FOR SURFACE SHIP WITH COMPENSATED TRACKING ERROR-CONSTRAINS AND DELAY INPUT BASED ON COMMAND-FILTER

Abstract

The article proposes an algorithm for trajectory tracking control problem of full actuated surface ships in the presence of state constraints, the delay of the input signal, and uncertain model parameters. During the design process, radial basis function neural networks are used to approximate the nonlinear components of uncertainty and a symmetric barrier Lyapunov function is incorporated to cope with the constraints of compensated tracking error. In particular, an auxiliary system is employed to eliminate the delay of the input signals, which often makes the control performance worse, even unstable. The adaptive controller that the article proposes is built based on the backstepping method using a command filter to avoid derivative explosion and reduce the computational burden on the controller. The article shows that tracking errors of surface ship can converge to a small neighborhood of zero, the compensated tracking error constraints of the system are not violated, the system is still stable when the input signal is delayed.