Fault detection of networked control systems based on fuzzy adaptive Interacting Multiple-Model algorithm

In this paper, the fault detection and the state estimation problems are investigated for a class of networked control systems with random transfer delay. First, Markovian jumping systems are modeled by assuming the random transfer delay satisfying a Markov chain. Then, Interacting Multiple-Model algorithm is used to detect fault and estimate system state at the same time based on the model developed. When the measurement covariance matrix changes abruptly, fuzzy logic is introduced to prevent the estimated error of Kalman filter occurring. Finally, a numerical example is presented to illustrate the effectiveness of the proposed algorithm.