Abstract
In this paper, we investigate the issues of state and fault estimation simultaneously for non-linear systems depicted by a Takagi-Sugeno (T-S) fuzzy models. The majority of current fault estimation schemes requires that the observer matching condition (OMC) should be fulfilled. Nevertheless, for many systems, it is not easy to achieve this condition. To surmount this limitation, a new fault estimation design is suggested, in which an intermediate variable is inserted and a suitably T-S fuzzy intermediate observer is suggested to estimate states and mismatched actuator faults simultaneously. The proposed concept is expressed and derived as a linear matrix inequality (LMI) problem. To confirm the effectiveness of this estimation scheme, a model of the longitudinal dynamics of an aircraft is introduced. Simulation results have shown that this approach successfully estimate states and actuator faults simultaneously despite the presence of mismatched faults.