基于液压泵复数信号的log-SAM故障诊断方法研究

针对基于液压泵单源振动信号易受噪声污染、无法有效表征故障特征信息的问题,提出了一种基于复数信号的log-SAM新方法。在液压泵端盖同一截面上,采集两个互相正交传感器上的振动信号;基于全矢思想,对上述双源信号融合为一个复数信号;利用log-SAM对复数信号进行包络分析,充分地突出故障特征信息,抑制噪声干扰;基于总体和最优诊断结果,实现故障诊断分析。通过分析实测液压泵滑靴磨损故障信号可知,所提方法具有很好诊断效果,且较传统全矢谱、基于单源信号分析方法和倒频谱预白化方法具有有效性和优越性。

Hydraulic pump fault diagnosis method using log-SAM on complex signals

A single signal from hydraulic pump is easily contaminated by noises,and the fault feature information is concealed.Thus,a new method using log-SAM based on complex signals was proposed.First,two orthogonal vibration signals were sampled on the same section of a pump end cover,and the two signals were fused into a complex signal based on the principle of full vector.Then,an envelop analysis on the complex signal was conducted by using log-SAM to extract clearly,the fault feature information and to filter out effectively the noises.Finally,the fault was diagnosed based on the comprehensive consideration of total fault diagnosis results and optimal fault diagnosis results.By virtue of an examplic testification by measured slipper wear fault signals processing,it is concluded that the proposed method is quite effective,and it performs better than the methods of the full vector spectrum,log-SAM based on a single signal and cepstrum pre-whitening.