Necessity for implementation of inverse problem theory in electric machine fault diagnosis

This paper presents the necessity of posing and solving inverse problems in the diagnostic procedures of electric machines. This is essential in order to achieve better and more precise results in the condition monitoring of the machines due to the growing complexity of the machine setups, which challenges the traditional electric machine diagnosis methods. An overview of the machine fault statistics as well as induction machine simulation results that put the credibility of the official statistics under question is presented. A method for using inverse problems in the machine diagnostic procedures is introduced and overview of existing problems solved via inverse problems is given. Necessity and directions of further study is pointed out.