Adaptive discrete-time fault diagnosis for a class of nonlinear systems: application to a mechanical manipulator

In this paper an adaptive approach to fault diagnosis (FD) for a class of nonlinear dynamical systems in the presence of actuator faults is presented. Namely, a discrete-time observer-based FD scheme is developed to compute an online estimate of the system state. In order to cope with uncertainties and discretization errors, a suitable parametric model is adopted for the uncertain terms, where the parameters are adaptively estimated. An experimental case study, based on an industrial mechanical manipulator, is developed to test the effectiveness of the approach.