Dynamic event-triggered adaptive neural prescribed performance control for dynamic positioning of vessels under input time-delay

Abstract

This paper presents a dynamic event-triggered adaptive neural prescribed performance control scheme for dynamic positioning systems of vessels in the presence of input time-delay and unknown time-varying ocean disturbances. First, a neural network is employed to compensate for the state time-delay term that is caused by the input time-delay. Second, the adaptive laws are designed to obtain estimates of the weights of neural networks as well as the bounds of the unknown time-varying ocean disturbances. By means of the prescribed performance error transformation, the dependency on priori knowledge of error initial values is overcome, and the stable transient and steady-state performance within the prescribed bounds are guaranteed. Then, a dynamic event-triggered mechanism is introduced to avoid unnecessary control operations and reduce the control burden. The backstepping design technique is utilized to derive the final control law such that the position and heading of vessels can be positioned on reference values under input time-delay and ocean disturbances. The highlights of this work lie in ensuring both high control performance and low control energy consumption simultaneously. The stability of closed-loop control systems is proved through the application of Lyapunov’s stability theory. Finally, the simulation results for two full-scale vessels validate the effectiveness.