Stability and stabilization of nonlinear systems and Takagi-Sugeno's fuzzy models (Q5950082)

This paper outlines a methodology to study the stability of Takagi-Sugeno fuzzy models. The stability analysis of the TS model is performed using a quadratic Lyapunov candidate function. This paper proposes a relaxation of Tanakas's stability condition: unlike related works, the equations to be solved are not Lyapunov equations for each rule matrix, but a convex combination of them. The coefficients of the associated sums depend on the membership functions. This method is applied to the design of continuous controllers for the TS model. Three different control structures are investigated, among which is the parallel distributed compensation. An application to the inverted pendulum is proposed here. Moreover, this new condition can be further applied to discrete fuzzy system and to the observation problem.