An adaptive observer for fault diagnosis in nonlinear discrete-time systems

This paper deals with the problem of fault diagnosis (FD) for a class of nonlinear systems in the presence of actuator failures. The scheme is based on a discrete-time diagnostic observer which computes an estimate of the system's state. To cope with the uncertainties and discretization errors, a discrete-time adaptive law is developed, based on a parametric model of the uncertain terms. A stability proof is developed to prove the global exponential stability of the system in the absence of faults. The effectiveness of the proposed approach is experimentally tested on a case study developed for an industrial mechanical manipulator.