A New Fuzzy Fault Diagnosis Method

An expanded type-2 fuzzy set is proposed, then we developed fuzzy diagnosis method based on it, in this paper; also the corresponding designing algorithm is given. As we added negative values to the membership function, the residual error outputs could be clearly divided into three statuses: normal, abnormal, fault. The added status could better "confirm evidence, eliminate feint". As a result, miss-diagnosis and misdiagnosis could be prevented. Aiming at a kind of nonlinear system, simulation result verifies the efficiency of this method.