The Research On Asynchronous Motor Fault Diagnosis Based On Artificial Intelligence Theory

Through diagnosing and analyzing the asynchronous motor faults essentially , the comprehensive evaluation of asynchronous motor operation status can be obtained,this contributes to find and locate faults, to make machine repair schedule, to stop the asynchronous motor from becoming worse and to reduce the loss from sudden accident. So this thesis carries out the asynchronous motor fault diagnosis research based on the artificial intelligence theory according to the asynchronous motor fault characteristics. Firstly, the basical asynchronous motor fault diagnosis principle is analyzed ,and its development , its situation at present and the characteristics and shortcomings of all kinds of diagnosis methods are expounded in this thesis. Secondly, both the artificial neural network and expert knowledge reasoning system are expounded , and the fault diagnosis system design thought based on the artificial neural network and expert knowledge reasoning is analyzed, and the expert system construction,the basical knowledge expression and acquisition methods, reasoning machanism and so on are introduced.In the end,the intelligent asynchronous motor fault diagnosis system combining information fusion with artificial network as well as expert knowledge system is presented. In this system, the fault signal acquisition and procession, the fault characteristics abstracting, the partial fault diagnosis and fusion policy are all included. The small wave analysis method is applied to abstract the fault characteristics in the message acquisition, and the parallel neuron network is used to initial diagnosis of the fault in the partial diagnosis, and the fusion method is applied to the fault reasoning procession according to the initial diagnosis results in the all-round diagnosis.Finally,the final diagnosis results are explained through expert rules.