Adaptive RBFNN-based fixed-time event-triggered control of a class of disturbed Euler-Lagrange systems with actuator faults

Abstract

This paper studies the radial basis function neural network (RBFNN)-based fixed-time event-triggered tracking control of a class of Euler-Lagrange (EL) systems with external disturbance, model uncertainties, and bias-actuator faults. A uniform robust exact differentiator (URED) is developed to estimate the accurate states of the EL systems within a constrained time. An adaptive RBFNN-based event-triggered control scheme based on the estimation signals are presented to inhibit the effects of the disturbances, uncertainties, and actuator faults. This scheme utilizes the minimum learning parameter (MLP) technique to significantly reduce both the update frequency of RBFNN parameters and the computational load, and also integrates the event-triggered mechanism to minimize controller updates, so that the resource utilization efficiency is enhanced. The stability of the EL error system under time constraints is proved using Lyapunov stability theory. Finally, simulations of a robotic manipulator system are demonstrated to display the effectiveness and superiority of the designed robust adaptive RBFNN-based fixed-time event-triggered fault-tolerant control strategy.