F. PourDadashi Komachali - Amirkabir University of Technology, Tehran, Iran.‎
M. Shafiee - Amirkabir University of Technology, Tehran, Iran.‎

The singular systems appear in many real occasions of system modeling. Fault occurrence is inevitable in real system; thus to avoid their destructive impacts, new design perspective must be taken. Performance and sensitivity of the fault diagnosis model based methods, however, significantly dependent on the accuracy of the model. In the one hand, it has been shown that many systems naturally follows the fractional order behavior, while on the other, in some scenarios, fractional modeling has improved the accuracy of the model. In this paper, we pay attention to the fault diagnosis in the fractional order singular systems. To this end, a singular observer with an unknown input has been used for diagnosis of the fault in the fractional order singular system, and the proposed observer convergence will be derived in the form of a linear matrix inequality. An advantage of the proposed method is separation of noise from the desired signal, both in inputs and outputs, using only the inputs and outputs signals.

Fault diagnosis, noise separation, singular system, fractional order system, linear matrix inequality.‎