Weak Fault Feature Extraction for Rolling Element Bearing Based on a Two-Stage Method

Timely and effective feature extraction is the key for fault diagnosis of rolling element bearing (REB). However, will become very difficult in early weak stage REB due to interference strong background noise. To solve above difficulty, a two-stage method proposed, which mainly combines mode decomposition (FMD) with blind deconvolution (BD) method. Firstly, based on impulsiveness cyclostationary characteristics vibration signal faulty REB, FMD used decompose complex original into several modes containing single component. Subsequently, sparse index (SI) calculated each mode, sensitive selected further analysis. apply enhancing impulsive characteristic. At last, traditional envelope spectrum (ES) analysis applied filtered signal, satisfactory features are extracted. Effectiveness advantages proposed verified through experimental engineering signals REBs.