State and Faults Estimation for T-S Models and Application to Fault Diagnosis

In this study, we propose a method to simultaneously estimate system state and faults for nonlinear systems described by T-S (Takagi-Sugeno) model. The method uses the technique of descriptor systems by considering both sensor and actuator faults as an auxiliary state variable. Then the idea consists in developing a descriptor observer for the given augmented system. This method has the advantage to estimate the state variables and the sensor and actuator faults simultaneously. The observer gains are determined using LMI (Linear Matrix Inequality) formulation. A three tanks system is used to show the effectiveness of the given result.