Adaptive Event-Triggered Asynchronous Fault Detection for Nonlinear Markov Jump Systems With Its Application: A Zonotopic Residual Evaluation Approach

In this paper, the event-triggered asynchronous fault detection (FD) problem is investigated for a class of nonlinear Markov jump systems (NMJSs) via zonotopic residual evaluation (ZRE). An adaptive event-triggered scheme (AETS) is utilized to reduce the redundant data releasing frequency such that both communication resources and network bandwidth can be saved as much as possible. The hidden Markov model (HMM) with generally hybrid probabilities is introduced into the FD framework to fully characterize the asynchronization phenomenon between the plant and the FD filter as well as the event generator and to address both the general transition and mode detection information issues simultaneously. By applying the double variables-based decoupling principle and variable substitution principle, a co-design criterion of optimal l₁/H asynchronous reduced-order FD filter and AETS is derived fully considering accessible mode information, such that the generated residual signal is sensitive to system faults while being robust to amplitude-bounded exogenous disturbances and measurement noise. Furthermore, a novel ZRE strategy with dynamic thresholds is developed for FD in NMJSs. Finally, an application of automotive electronic throttle body is used to illustrate the effectiveness of the proposed approach.