Fault Diagnosis For Rolling Bearing With Variable Speeds Based On EMD Re-sampling And Rough Sets

Machinery in a changing environment has amounts of non-stationary signals within its vibration signal. Majority of theories of fault diagnosis are based on the characteristics that reflect on the stationary speed machine. However, with the development of machinery and the need of the technology, how to deal with non-stationary signals is becoming a key issue in the field of fault diagnosis.On account of rolling bearing with variable speeds, this article puts forwards an effective method with EMD resampling and rough sets. The EMD resampling method can make the IMF of the bearing with variable speeds' signal stationary in frequency domain. The method solved the problem caused by the changing speeds. After resampling, the signal is easy to use for fault diagnosis with the rough sets.First of all, the article analyzed the superiority of EMD to deal with non-stationary signals and the process the signal of variable speeds with the EMD resampling method. As a result, a pretty method and the new characteristics of IMFs stand out.Second, the article analyzed the sensibility of the traditional characteristic parameters in time domain and frequency domain of the results of the EMD resampling. At last, we got some characteristic parameters with noticeable features.Finally, the article analyzed the rough sets method and make use of its advantage to deal with the characteristics of the EMD resampling signals. We finally got the decision table for fault diagnosis. We also proved it an effectively method of the fault diagnosis of the rolling bearing with variable speeds according to the data we got in experiment.