Abstract: Rotor crack is a very dangerous and frequent fault in rotating machine, and crack detection and location is a always research direction in the field of fault diagnosis. The crack detection as a inverse problem and propose genetic algorithms based rotor crack detection strategy is regarded, which formulates the rotor crack detection as an optimization problem by means of finite element method and utilizes genetic algorithms to search the solution. The numerical experiments suggest that good predictions of the rotor crack locations and configuration are possible and the proposed strategy is feasible. The study also indicates that the proposed method has the potential to solve a wide range of inverse identification problems, such as non-destruction detection and flaw identification, in a systematic and robust way.

Key words: Finite element method, Genetic algorithms, Inverse problem, Rotor crack