Robust adaptive FTC allocation for over‐actuated systems with uncertainties and unknown actuator non‐linearity

Abstract

In this study, a robust adaptive fault-tolerant control (FTC) allocation method is proposed to address the FTC problem of over-actuated systems in the presence of matched disturbance, unmodelled dynamics and unknown actuator non-linearity simultaneously. The main idea is to design the virtual control law via robust adaptive control and then distribute the virtual control among individual actuators by weighted pseudo-inverse control allocation whether the actuators are faulty or not. The virtual control consists of three terms: a linear term is designed through adaptive control to maintain stability, a non-linear term constructed by radial basis function neural network (RBFNN) is used to approximate the unmodelled dynamics, and a robust term is added to eliminate the approximation error introduced by RBFNN as well as matched disturbance and unknown actuator non-linearity. With the aid of the Lyapunov stability theorem, the convergence of the closed-loop system is proven. The simulation results demonstrate the effectiveness, fault-tolerant capability and robustness of the proposed method.