Motor Online Fault Diagnosis Based on Artificial Intelligence Techniques
Chidong Qiu, Yue Tan, Guang Sheng Ren · 2006
This paper introduces a hybrid method of motor online fault diagnosis. Firstly, motor stator current was processed by means of fast Fourier transform, the frequency response was gained. Then the continuous frequency response was discretized for the aim of classification, the discretization process was implemented based on Kohonen neural networks. Finally, the advanced classification was implemented by means of rough set theory. Based on reduced decision table, fault diagnosis rules were found. In this paper, those classified data was measured in laboratory when motor operated under man-made fault condition. By simulating and computing, it was confirmed that the method was feasible