基于emd与ica的滚动轴承复合故障诊断-北京工业大学学报.pdfVIP

第40卷 第10期 北 京 工 业 大 学 学 报 Vol.40 No.10 2014年 10月 JOURNAL OF BEIJING UNIVERSITY OFTECHNOLOGY Oct. 2014 基于EMD 与ICA 的滚动轴承复合故障诊断 崔玲丽,吴春光,邬摇 娜 (北京工业大学 机械工程与应用电子技术学院先进制造技术北京市重点实验室,北京摇 100124) 摘摇 要:针对单通道情况下滚动轴承复合故障难以分离问题,提出基于经验模式分解 (empirical mode decomposition,EMD)的独立分量分析(independent component analysis,ICA)算法. 该方法首先对单通道采集的轴承 复合故障信号进行EMD分解,得到多个基本模式分量函数(intrinsicmodefunction,IMF),然后依据帩度指标及相关 系数值,选取有效的IMF分量与原观测信号组成新的观测信号,对其进行ICA处理,进而实现轴承复合故障的分 离. 实验结果表明,该方法可有效地分离轴承早期的复合故障. 关键词:经验模式分解;独立分量分析;峭度指标;滚动轴承复合故障诊断 中图分类号:TH16 文献标志码:A 文章编号:0254-0037(2014)10-1459-06 Composite Fault Diagnosis of Rolling Bearings Based on EMD and ICA Algorithm CUI Ling鄄li,WU Chun鄄guang,WU Na (The Key Laboratory of Advanced Manufacturing Technology,College of Mechanical Engineering and Applied ElectronicsTechnology,Beijing University of Technology,Beijing 100124,China) Abstract:Aiming at the difficulty of separating rolling bearing composite faults in the case of single channel signal, a new algorithm based on empirical mode decomposition (EMD) and independent component analysis (ICA) was proposed in this paper. First, the composite bearing fault signals collected for a single channel were decomposed by EMD to obtain some intrinsic mode function (IMF). Then,main components were confirmed by calculating the correlation coefficient of every IMF and original composite signal and kurtosis value of every IMF. At the same time, main components with original signal were processed by ICA to realize the separation of composite fault of rolling bearings. Experimental results show that thismethod can effectively separate early rolling bearingscompositefau