Bearing Fault Diagnosis Method Based On Emd And Resonance Demodulation Study

The rolling element bearing is one of the most important parts in rotation machineries, and plays an important role in functionality and performance. Any flaw in bearings might results in the malfunction or even catastrophic failure of the equipment, so it is of great significance to evaluate the performance of the bearings with out interrupting the operation of the machineries.In this paper, the mechanical structure, the vibration principle and the fault categories of the rolling element bearing were presented. And the approach to defect the fault categories, such as the resonance demodulation, Hilbert transform and statistics, were also exploited. Further, we also designed a data acquisition system to collect the vibration signal of the accelerated lifetime intensifying test rig.With the research about resonance demodulation and Hilbert-Huang transform find that diagnosis based resonance demodulation has weaknesses that need to pre-determined and fixed the parameters of band pass filter such as center frequency and filter band width, and diagnosis by Hilbert-Huang transform is able to describe signal change both time scale and frequency, but it is difficult to recognize fault characteristic of signal by Hilbert spectrum and marginal spectrum. The paper proposed combination of empirical mode decomposition and the technique of resonance demodulation to diagnosis fault . Adaptability of EMD fill the gap need to fixed the parameters of band pass filter , and the technique of resonance demodulation fill the gap that HHT can't obviously describe fault characteristic by extract fault information from high-frequency vibratory signal. The experiment shows the combination of the EMD and Hilbert transform is effectiveness and defect the different fault categories successfully. In EMD resonance demodulation diagnostic method, there are shortcomings of producing many false IMF components and selecting IMF components for feature extraction required manual intervention. An improved EMD resonance demodulation fault diagnostic method is proposed in this paper. It's compared and analyzed with both traditional resonance demodulation diagnostic method and EMD resonance demodulation diagnostic method in computation load, intelligent degree and valid aspect. The analysis result indicate that improved EMD resonance demodulation fault diagnostic method which proposed in this paper is superior to traditional resonance demodulation method and EMD resonance demodulation method.