关键词: 滚动轴承, 特征提取, 微弱故障, 稀疏分解, 最小熵解卷积

Abstract: The rolling bearing is one of the key mechanical parts whose fault diagnosis is very important. The rolling bearing’s fault feature under strong background noise is very weak for reasons of environment noise impact and signal attenuation. The feature extraction of rolling bearing’s weak fault is not only very important but also very hard. The sparse decomposition has been used in the fault feature extraction of rolling bearing. But its performance is very poor when the background noise is very strong. The minimum entropy deconvolution (MED) and sparse decomposition are combined for rolling bearing’s weak fault diagnosis. The strong background noise of rolling bearing is decreased by the MED method firstly, then the de-noised signal is handled by the sparse decomposition. At last the envelope demodulation is carried on the last given signal and better result is obtained. In the end through simulation signal and experiment the effectiveness and advantage of the proposed method are verified.

Key words: Feature extraction, Minimum entropy deconvolution, Rolling bearings, Sparse decomposition, Weak fault