基于形态小波和S变换的滚动轴承故障特征提取

针对传统小波在强背景噪声中提取冲击故障特征的不足,提出基于极大提升形态小波(MLMW)分析和S变换的滚动轴承故障特征提取方法.先利用MLMW变换将信号分解到不同形态尺度上,各尺度信号上保留着信号局部极值形态特征,对细节信号进行软阈值降噪处理,再从重构信号的具有良好时频聚焦性的S变换谱上提取故障特征.试验结果表明,MLMW既抑制了噪声和谐波分量,又显著强化了故障特征;相比传统小波和包络分析,能清晰地提取非平稳非线性故障特征.由于MLMW采用简单的形态算子和高效的提升方法,计算简单高效,适于故障特征的在线分析.

Rolling bearing fault feature extraction based on morphological wavelet and S-transform

Aimed at the deficiency of traditional wavelet extracting impulse fault features from strong noise background, a fault feature extraction method for rolling bearing was proposed based on max-lifting morphological wavelet(MLMW) and S-transform. Firstly, decomposed to different levels by MLMW, signal’s local maxima were mapped to scale signals and preserved over several scales, and the detail signals were denoised by soft threshold. Secondly, the signal was reconstructed, and fault features were extracted from the denoised signal‘s S-transform spectrum with excellent time-frequency focus characteristic. The experimental results show MLMW analysis not only reduces noise and harmonic components, but also significantly enhances fault features, and can extract nonlinear and non-stationary fault features more clearly than classical wavelet transform and envelopment analysis. Furthermore, with simple morphological operators and efficient lifting scheme adopted, the MLMW algorithm is simple and the cost is low, so it is suitable for on-line fault features analysis．