LCD全矢包络谱及其在轴承故障特征提取中的应用

摘要

针对电机滚动轴承故障特征提取问题,提出了一种基于局部特征尺度分解(Local Characteristic Decomposition,LCD)的全矢包络谱(Full Vector Envelope Spectrum,FVES)技术.该方法通过正交采样技术获取两个互相垂直方向的振动信号,然后用LCD分别将其分别分解成系列内禀模态分量(Intrinsic Scale Component,ISC)之和;其次,用Hilbert变换对两个方向的ISC进行包络解调得到包络信号;最后运用FVES融合两个方向上的包络信号得到全矢包络谱.电机滚动轴承外圈故障分析结果表明,较之单一方向上信号的包络谱,全矢包络谱的频谱结构更为清晰、幅值更大.%In order to diagnose AM-FM class of motor failure,a full vector envelope spectrum technology (FVES)based on local characteristic decomposition (LCD) was proposed.Two directions of vibration signals were got by orthogonal sampling.Then,the signals were decomposed by LCD respectively to obtain intrinsic mode components of the series (ISC).Secondly,envelope signals of ISCs were obtained by Hilbert Transform.Full vector envelope spectrum was obtained by vector spectrum technology.Rolling bearing outer ring failure analysis results show that full vector spectrum envelope spectrum structure and amplitude is clearer and greater,compared with a single direction signal envelope spectrum.