Switching Adaptive Compensation for Actuator Failures of Uncertain Nonlinear Systems with Quantized Input

Abstract

This paper investigates the adaptive compensation for an uncertain nonlinear system with actuator failures and input quantization. A new switching fault-tolerant controller is proposed based on the dwell time bounded by the minimal time length before total loss of effectiveness (TLOE) of an actuator group and the maximal time length spending for an actuator group recovering from TLOE mode. By estimating the infimum of the fault parameters for all active times, the fault for each actuator is allowed to appear infinite many times. The boundedness of the (adjustable) jump size of Lyapunov functions for different actuator groups at switching instants is guaranteed by exploiting the parameter projection operator. Under the designed controller and switching signal, the output tracking error of the system can be driven to a tunable neighborhood of the origin. The proposed method is validated by a numerical example at last.