基于数学形态学的水泵振动故障诊断研究

Diagnostic Research on Vibration the Faults of Pumps Based on Mathematical Morphology

基于多尺度数学形态谱及形态谱熵等理论,对多组实测水泵故障信号分析对比其形态谱特征和形态谱熵,提取故障信号特征,并结合信号时域图,达到故障识别的目的。研究表明:采用数学形态谱和形态谱熵理论能够对杂乱无章的故障进行特征提取,故障信号识别成功率在90%以上,较好地达到了故障诊断的目的,这为信号处理和水泵故障诊断提供了新方法,具有较高的实用价值和学术价值。

Based on the multi-scale mathematical morphology spectrum and morphological spectrum entropy, the tested pump vibra- tion signal is analyzed to extrude the features of signals. And by combining the time domain graph of signals, the morphology spec- trum as well as corresponding entropy is analyzed to identify faults accurately. The research shows that the method of morphology spectrum and morphology spectrum entropy can effectively extrude the features of signals, and accuracy is more than 90%, so it is an effective method in signal processing and the fault diagnosis of pumps, and has a high application value and academic value.