Neural Network-Based Adaptive Fault-Tolerant Control for a Class of 1-D System with Actuator Failures and Unknown Input Powers

Abstract

This paper studies fault-tolerant control problem for a class of 1-D nonlinear systems with actuator failures and unknown powers. On the basis of Lyapunov stability theory, an adaptive control strategy is put forward and adopted for offsetting actuator failures to ensure the system’s output can follow reference signal. In existing works, actuator failures occur in nonlinear systems have been addressed. However, the input powers of these systems, especially the powers of controllers are either limited to one or a known positive integer. To relax such restrictions, we consider more general situation where controllers’ input powers are extended to an unknown integer greater than one and both efficiency loss and bias occur in the above actuators. Finally, a numerical example demonstrates the general effectiveness of developed algorithm.