基于约束独立成分分析的轴承复合故障特征提取方法.pdfVIP

第35 卷 第3 期 噪 声 与 振 动 控 制 Vol 35 No.3 2015 年6 月 NOISE AND VIBRATION CONTROL Jun. 2015 文章编号：1006-1355(2015)03-0173-04 基于约束独立成分分析的轴承复合故障特征 提取方法 李瑞彤，王华庆，屈红伟，齐 放，李美娇 （北京化工大学 机电工程学院，北京 100029 ） 摘 要：为从复合故障信号中提取各故障特征，提出一种离散小波变换（DWT ）和约束独立成分分析（CICA）相结合 的单通道复合故障诊断方法。首先通过DWT 方法将单通道振动信号进行小波分解后，利用小波重构函数重构各层分 解信号。然后取重构信号的包络信号作为CICA 算法的输入矩阵，基于滚动轴承先验知识建立参考信号，从而分离出 轴承各故障信号，提取故障特征。最后，在滚动轴承故障模拟实验台上进行了方法验证。结果表明：该方法可有效分 离滚动轴承外圈和滚动体故障，实现了轴承复合故障的诊断。 关键词：振动与波；复合故障诊断；约束独立成分分析；离散小波变换；滚动轴承 + +.3 中图分类号：TH 133.3 ；TH 165 .3 ；TP206 文献标识码：A DOI 编码：10.3969/j.issn. 1006-1335.2015.03.037 Application of CICA in Compound Fault Feature Extracting of Rolling Bearings LI Rui-tong , WANG Hua-qing , QU Hong-wei, QI Fang , LI Mei-jiao ( College of Mechanical and Electrical Engineering, Beijing University of Chemical Technology, Beijing 100029, China ) Abstract : In order to extract fault features from compound signals, a method based on discrete wavelet transform (DWT) and constrained independent component analysis (CICA) was proposed. In this method, the single channel vibration signal was decomposed into several wavelet coefficients by DWT method, and the wavelet re-construction function was used to reconstruct the decomposed signal. Then, envelope signals of the reconstructed wavelet coefficients were selected as the input matrix of CICA algorithm, and the reference signal was established based on prior knowledge of source signals. Finally, the fault signals were separated and the fault features