非平稳机械振动噪声中瞬态故障信号的检测  

Extraction of instantaneous-fault signal in unsteady mechanical vibration signal

作　　者：鲁怀伟[1] 马莉[1] 蒲会兰[1] 罗冠炜[2] 

机构地区：[1]兰州铁道学院基础科学系,甘肃兰州730070 [2]兰州铁道学院机电与动力工程学院,甘肃兰州730070

出　　处：《铁道学报》2003年第1期68-72,共5页Journal of the China Railway Society

基　　金：国家自然科学基金资助项目(10172042);铁道部科技专项基金(J2000Z082)

摘　　要：结合非平稳机械运行状态,分析了非平稳振动信号中瞬态故障信号和平稳振动信号的特点,介绍了应用小波包理论进行故障信号检测的基本原理.通过小波包变换后,平稳振动信号和瞬态故障信号的小波包-能量谱表现出不同的特性.本文运用统计信号处理的理论,构造了相应统计量进行假设检验判决,判断平稳振动信号中是否存在瞬态故障信号。针对变速箱故障声压信号的非平稳时变特点,给出了小波包分析处理变速箱故障声压信号的原理和方法,并分析了一个变速箱齿根裂纹故障的诊断实例.实验结果表明,该方法可以有效地提取淹没在平稳振动噪声信号中的瞬态故障信号的特征,准确判断出故障,验证了该方法对机械设备进行故障诊断的有效性,对其它类型的机械故障诊断也具有一定的参考价值。Considering the condition of unsteady vibration of machinery, and based on the analysis of characteristic of fault signals generated by mechanisms and steady vibration signal, the basic theory of wavelet packet decomposition is proposed in this paper. Aiming at the characteristic of unsteady and timevariation of gearbox's acoustic signal, the wavelet packet theory is introduced into gearbox fault diagnosis. An example of acoustic signal processing and mechanical fault diagnosis via wavelet packet analysis is provided. The result demonstrates that the method is good for acoustic signal processing and fault diagnosing in gearbox. At the same time, an effective method is provided for diagnosing the working state of all kinds of mechanisms such as internalcombustion, diesel engine and so on. 

关 键 词：故障诊断 小波包 信号滤波 非平稳机械振动 

分 类 号：TH165.3[机械工程—机械制造及自动化]