S. Pöyhönen, P. Jover, and H. Hyötyniemi (Finland)
Electrical Machines, Vibration Monitoring, Fault Classification, Data Fusion, Independent Component Analysis, Support Vector Machine
ICA is applied to multi-channel vibration measurements of a 35 kW cage induction motor to fuse the information of several channels, and provide a robust and reliable fault detection routine. Independent components are found from the measurement data set with FastICA algorithm, and their PSD estimates are calculated with Welch's method. A SVM based classification routine is applied to the PSD estimates to perform the fault identification. Similar classification is applied directly to vibration measurements. Based on the results with real measurement data it is shown that data fusion with ICA enhances the fault diagnostics routine.
Important Links:
Go Back
