IMF分量的倒频谱分析在滚动轴承故障诊断中的应用

轴承局部损伤故障引起的振动响应往往会被较大的振动信号所掩盖,影响故障的正确诊断。本文提出IMF分量的倒频谱分析方法,首先将复杂的信号分解为有限的内禀模态函数（IMF）之和,对原始振动信号进行降维;再对IMF分量进行倒频谱分析,利用倒频谱方法准确地提取振动信号幅值谱上的周期特征。对EMD分解得到的多个IMF分量同时做倒频谱分析,可以相互验证,从而得到更可靠、更准确、更可信的诊断结果。通过对IMF分量的倒频谱分析法和Hilbert包络谱分析法诊断效果进行比较,结果表明,IMF分量的倒频谱分析方法比Hilbert包络谱分析提取的故障频率特征更精准、可靠。

Cepstrum Analysis of the IMF Components in Ball Bearing Fault Dingnosis

Bearing local damage caused by vibration failures tend to be overshadowed by the larger vibration signals affect the correct fault diagnosis.A method based on amplitude IMF cepstrum analysis for bearing fault diagnosis was introduced.First,the complex signal is decomposed into a sum of the limited intrinsic mode function(IMF),and to dimensionality reduce the original vibration signal.Then doing cepstrum analysis on the component of the IMF to extract vibration signal amplitude spectrum of the cycle characteristics.Simultaneously doing cepstrum analysis on a number of IMF components can be mutual authentication.Thus obtain more reliable and accurate diagnosis.The experimental results show that the Failure frequency characteristics obtained by the cepstrum analysis of IMF component is more accurate and reliable.