Detection of induction motor faults by an improved artificial ant clustering

In the last decade, the field of diagnosis has attracted the attention of many researchers, especially for the diagnosis of induction motors. This type of machine is widely used in industry because of its robustness and its specific power. Therefore, the monitoring and diagnosis of these motors become very important. This paper deals with the diagnosis of induction motor faults. The method is based on ant-clustering and it is improved by K-means pattern recognition and Principal Components Analysis (PCA). This approach is applied to the diagnosis of a squirrel-cage induction motor of 5.5kW with broken bars and bearing faults in order to check the detection capability. The obtained results prove the efficiency of this approach.