Fault detection for discrete-time Markov jump systems with time-varying delays and partially known transition probabilities

This paper investigates a fault detection (FD) method for discrete-time Markov jump systems with time-varying delays and partially known transition probabilities. Different from the traditional systems that need to know all transition probabilities, the proposed method only needs to know the partially known transition probabilities to realize FD, which relax the condition greatly. The FD design is reduced to a ${{\\text{H}}_\\infty }$ filtering problem through the FD filter is regarded as residual generator. A numerical example demonstrates effectiveness of the proposed method.