Observer based fuzzy-control design using relaxed LMI conditions

This paper proposes two different approaches to design observer based regulators for T-S fuzzy models. The first approach has a discrete-time state feedback counterpart in the literature. Here, we derive appropriate LMI conditions for the continuous-time case. Furthermore, new LMI conditions are derived to facilitate the design of an observer- based regulator. The second approach generalizes the first one by using a new parallel distributed compensator with a modified quadratic Lyapunov function. By including constraints on the input signal, it is shown that the second approach can cope with tighter bounds on the control signal. A typical simulation example is included to illustrate the performance of both design approaches.