Reliable LQ fuzzy control for nonlinear discrete-time systems via LMIs.

Huai-Ning Wu
DOI: https://doi.org/10.1109/TSMCB.2003.817038
2004-01-01
Abstract:This paper studies reliable linear quadratic (LQ) fuzzy regulator problem for nonlinear discrete-time systems with actuator faults. The Takagi and Sugeno fuzzy model is employed to represent a nonlinear system. A sufficient condition expressed in linear matrix inequality (LMI) terms for the existence of reliable guaranteed cost (GC) fuzzy controllers is obtained. The fuzzy controller directly obtained from the LMI solutions can guarantee the stability of the closed-loop overall fuzzy system, while provide a guaranteed cost on the quadratic cost function of the system in the normal and actuator fault cases. Furthermore, an optimal reliable GC fuzzy controller in the sense of minimizing a bound on the worst or nominal case guaranteed cost is also given by means of an LMI optimization procedure. Finally, a numerical example is provided to demonstrate the effectiveness of the proposed method.
What problem does this paper attempt to address?