DVSL guidance-based composite neural path following control for underactuated cable-laying vessels using event-triggered inputs

Abstract

In this note, the path following issue of underactuated cable-laying vessels is addressed from two aspects, i.e., guidance and control. By exploiting the hierarchical design, a so-called dual virtual ships layered (DVSL) guidance algorithm is proposed with two reference path generators. In the first guidance layer, the logical virtual ship (LVS) is used to programme the setpoints-based reference path for cables. Based on the cable-related geometrical derivation, the corresponding ship’s reference path can be obtained in the second layer, and its smoothness can be guaranteed by the adaptive virtual ship (AVS). For the control part, a novel composite neural event-triggered control algorithm is developed by taking the main engine revolution and the rudder as actual control inputs, which are characterized by the practical measurability and the aperiodic update. The radial basis function neural networks (RBF-NNs) are employed to deal with the model uncertainty, and the approximation performance is improved with the aid of the serial-parallel estimation model (SPEM). By fusing the robust neural damping technique, only two adaptive parameters require to be updated, which leads to a simpler controller with less computation burden. Besides, the inputs-dominated event-triggered mechanism is introduced to release the communication traffic in the controller-to-actuator channel. Through the Lyapunov stability analysis, all signals in the closed-loop system are guaranteed to be semi-global uniformly ultimately bounded (SGUUB), and the existence of minimum inter-event time is proved. Finally, three simulation experiments are conducted to illustrate the effectiveness and the superiority of proposed algorithm.