A model based fault detection and accommodation scheme for nonlinear discrete-time systems with asymptotic stability guarantee

In this paper, a fault detection and accommodation (FDA) framework is developed for unknown nonlinear discrete-time systems. The changes in the system dynamics due to the faults are modeled as a nonlinear function of state and input variables while the time profile of the fault is assumed to be exponentially developing. A fault is detected by monitoring the system states and reconstructing the fault dynamics using online approximators. The online approximator output is used first for fault detection and later reconfigured for accommodation. A stable adaptation law in discrete-time is developed not only to characterize the faults but also for controller reconfiguration. The asymptotic stability of the closed-loop system due to the FDA algorithm is demonstrated in the presence of online approximator reconstruction errors and bounded system uncertainties by using a robust term. Finally, a simulation example is utilized to illustrate the performance of the proposed FDA scheme.