| Summary: | 碩士 === 國立中興大學 === 機械工程學系所 === 104 === In this research, the envelope analysis and relative period spectral analysis are utilized to diagnose the different defective classes of bearing under different stationary speed, including the single defect on inner race, outer race, and roller as well as the double defects on outer race. First, the empirical mode decomposition is employed to decompose the signals into a number of intrinsic mode functions. The intrinsic mode functions that have obvious amplitude modulation phenomenon are then selected for envelope analysis. The successional procedure consists of two parts. In the first part, the fast Fourier transform algorithm is used to extract the frequency-domain features for identifying the single-defect classes. In the second part, the relative period spectrum is employed to extract the periodic features for diagnosing the double-defect classes. Finally, the support vector machine is utilized to classify the classes of signle-defect as well as the double-defect.
|
|---|
