Model free adaptive fault-tolerant tracking control for a class of discrete-time systems

Abstract

This paper investigates the fault-tolerant tracking control issue for a class of discrete-time systems with sensor fault. A novel model-free adaptive fault-tolerant tracking control scheme is proposed by utilizing the dynamic linearization approach, in which only input/output (I/O) data of controlled plant are required. Moreover, due to the characteristics of simple structure and powerful approximation ability, RBF neural network (RBFNN) is introduced to approximate the fault dynamics after the sensor fault is detected. The approximated fault dynamic is utilized to reconstruct the controller. The major feature of this paper is that the whole fault-tolerant tracking controller structure is derived from the dynamic linearization of the ideal controller, rather than the criterion function. Finally, the proposed model-free adaptive fault-tolerant tracking controller is verified to be effective by numerical simulations.