Adaptive fault diagnosis for robot manipulators with multiple actuator and sensor faults

Abstract

This paper investigates the fault detection and isolation (FDI) problem for robot manipulators with actuator and sensor faults. The considered manipulators are modeled by a class of nonlinear systems with Lipschitz-like nonlinearities and modeling uncertainties. The simultaneous occurrence of actuator and sensor faults is considered, which results in that feedback information for FDI is corrupted and the residual signals might be sensitive to both actuator faults and sensor faults. Under a mild persistent excitation condition, a nonlinear adaptive observers is constructed to exponentially converge with a pre-specified estimation error bound. The performances of the proposed FDI scheme, including robustness of fault estimate to the uncertainties, accuracy of fault estimation and rapidity of fault diagnosis, are rigorously analyzed.