| In this paper Support Vector Machines (SVM) was used in fault intelligent diagnosis of machine element such as gear and bearing. The research was included:Feature extraction: The feature value of time domain includes peak value, peak to peak value, kurtosis and so on. The feature value of frequency domain is MSF. The SVM method was used for detecting the rolling bearing fault of fan in power plant.From the vibration signals of reference and fault stations the feature of time domain and the feature of frequent domain. The results showed that the reference and fault stations of fan can be distinguished clearly in the SVM diagram.The SVM method was also used for detecting the gear case. The feature of time and the feature of frequent was also be used. The results showed that the SVM can distinguish the different fault in different time.Through designed a band-pass filter, the feature of gear case's signal was extracted, including feature of time and feature of frequent. The results showed that it was better than that signals which didn't use filter. |
