H∞ fuzzy control design of discrete-time nonlinear active fault-tolerant control systems

This paper is concerned with the problem of H-infinity fuzzy controller synthesis for a class of discrete-time nonlinear active fault-tolerant control systems (AFTCSs) in a stochastic setting. The Takagi and Sugeno (T-S) fuzzy model is employed to exactly represent a nonlinear AFTCS. For this AFTCS, two random processes with Markovian transition characteristics are introduced to model the failure process of system components and the fault detection and isolation (FDI) decision process used to reconfigure the control law, respectively. The random behavior of the FDI process is conditioned on the state of the failure process. A non-parallel distributed compensation (non-PDC) scheme is adopted for the design of the fault-tolerant control laws. The resulting closed-loop fuzzy system is the one with two Markovian jump parameters. Based oil a stochastic fuzzy Lyapunov function (FLF), Sufficient conditions for the stochastic stability and H-infinity disturbance attenuation of the closed-loop fuzzy system are first derived. A linear matrix inequality (LMI) approach to the fuzzy control design is then developed. Moreover, a suboptimal fault-tolerant H-infinity fuzzy controller is given in the sense of minimizing the level of disturbance attenuation. Finally, a simulation example is presented to illustrate the effectiveness of the proposed design method. Copyright (c) 2008 John Wiley & Soils, Ltd.