Search

This paper is concerned with the fault detection problem for a class of networked fuzzy systems with multiplicative noises on both system states and measurement outputs. In view of the limited communication capacity, a periodical communication protocol (i.e., the round-robin protocol) is adopted to undertake the transmission task between the sensors and the fault detection filter, which leads to periodical delays in the overall system. A Takagi–Sugeno fuzzy-model-based fault detection filter is constructed to produce the residual signal and an auxiliary error system is established to facilitate the stability analysis of the error dynamics. With the aid of Lyapunov stability theory, sufficient conditions are obtained that ensure the exponentially mean-square stability of the error dynamics with prescribed $H_{\infty }$ performance constraint. The desired fault detection filter is designed by solving a convex optimization problem via the semidefinite programme method. A finite-time evaluation function and an adjustable threshold are introduced in order to detect the possible faults effectively. The effectiveness of the proposed fault detection scheme is validated by a numerical simulation example. [ABSTRACT FROM AUTHOR]

This paper is concerned with the robust H{\infty } fault detection problem for a class of uncertain discrete-time nonlinear 2-D systems subject to Round-Robin scheduling protocol. The Takagi–Sugeno fuzzy model is used to approximate the nonlinearities, where the linear fractional uncertainties enter the system in a random way. A kind of widely used communication mechanism, namely, Round-Robin communication protocol, is adopted to periodically schedule the sensors and the fault detectors to realize the information exchange in order to reduce the bandwidth usage in a networked environment with limit resource. An improved 2-D fuzzy residual generator is constructed to detect the possible fault, where the stability analysis of the resulting augmented 2-D system is discussed. It is accomplished by using a combination of the basis-dependent Lyapunov-like function and the stochastic analysis techniques. Sufficient conditions are first established to guarantee the globally asymptotic stability of the error dynamics of the state estimation with prescribed H{\infty } performance constraints. Then, a residual generator is proposed to detect the possible faults. The effectiveness of the developed algorithm is demonstrated via application to the fault detection problem for a thermal process. [ABSTRACT FROM AUTHOR]