Nonlinear ELM Estimator-Based Formation Control for Perturbed Unmanned Surface Vehicle Systems with Prescribed Performance

This paper addresses the challenge of achieving resilient formation control for a class of multiple unmanned surface vehicles (USVs) systems operating amidst perturbations, which encompass external disturbances and parametric uncertainties. In order to alleviate the unfavorable effects of perturbations, an innovative nonlinear perturbation estimator rooted in the extreme learning machine (ELM) is introduced to meticulously evaluate the perturbations. Based on the estimation results, a perturbation-rejection formation control scheme is designed using dynamic surface control (DSC) and prescribed performance control (PPC) technologies. The aim is to prevent complexity explosion and ensure that the errors of the multiple USVs system converge to predefined regions, thereby enabling USVs to avoid collisions and prevent communication interruptions. The asymptotic formation control of the multiple USVs system is substantiated through Lyapunov stability theory. Ultimately, the effectiveness of the proposed perturbation-rejection formation control technique, based on the nonlinear ELM-based estimator, is validated through simulations.