基于参考独立分量分析的頻混信号分离研究
详细信息 查看官网全文
- 英文论文题名：Study on Diagnosis of the Mixed Faults Based on Constrained Independent Component Analysis
- 论文作者：王志阳 ; 陈兰 ; 张永鑫 ; 周红梅
- 英文论文作者：WANG Zhi-yang ; CHEN Lan ; ZHANG Yong-xin ; ZHOU Hong-mei ; School of Mechanics & Power Engineering ; Henan Polytechnic University
- 年：2016
- 作者机构：河南理工大学机械与动力工程学院;
- 论文关键词：独立成分分析 ; 约束独立成分分析 ; 盲源分离 ; 机械故障诊断 ; 混合故障诊断
- 英文论文关键词：Independent component analysis ; Constrained independent component analysis ; Blind source separation ; Machine fault diagnosis ; Mixed faults diagnosis
- 会议召开时间：2016-07-28
- 会议录名称：第二十七届全国振动与噪声应用学术会议论文集
- 语种：中文
- 分类号：TH17
- 学会代码：ZDZD
- 会议名称：第二十七届全国振动与噪声应用学术会议
- 会议地点：中国黑龙江哈尔滨
- 主办单位：中国振动工程学会振动与噪声控制专业委员会
- 学会名称：中国振动工程学会振动与噪声控制专业委员会
- 页数：4
- 文件大小：10756k
- 原文格式：D
- 会议级别：全国
摘要
独立成分分析(ICA)是利用数据的统计独立性进行源分离的信号处理方法,它具有幅值不定性和顺序不定性。约束独立成分分析方法(CICA)把诊断对象的先验信息作为约束条件,使ICA算法仅仅收敛于感兴趣的信号,这样可以提高故障诊断的针对性和有效性。这种方法可以快速诊断传感器信号中是否有某种故障。本文设计了频率重叠的混合故障,用约束独立成分分析方法和平方包络法进行对比研究。仿真和实验验证了该方法的在混合故障诊断中适用性和有效性。
The mixed signals can be separated by independent component analysis(ICA)using statistics of the data.It is sometime invalid in machine fault diagnosis for amplitude and order uncertainty from ICA.The prior information of the object as a constraint is integrated into the independent component analysis algorithm based on constrained dependent component analysis,the new algorithm converges only fault signal of interest in the paper and is a fast diagnosis approach for the specific faulty type in the machine fault diagnosis.The relevance and effectiveness of fault diagnosis are improved.The method can find out rapidly whether there is a fault in mechanical equipment from sensor signals.The applicability and effectiveness of the method are verified by the simulation and experiment of the mixed faults whose frequencies are superimposed.
The mixed signals can be separated by independent component analysis(ICA)using statistics of the data.It is sometime invalid in machine fault diagnosis for amplitude and order uncertainty from ICA.The prior information of the object as a constraint is integrated into the independent component analysis algorithm based on constrained dependent component analysis,the new algorithm converges only fault signal of interest in the paper and is a fast diagnosis approach for the specific faulty type in the machine fault diagnosis.The relevance and effectiveness of fault diagnosis are improved.The method can find out rapidly whether there is a fault in mechanical equipment from sensor signals.The applicability and effectiveness of the method are verified by the simulation and experiment of the mixed faults whose frequencies are superimposed.
引文
[[1]Long Han,Cheng Wei Li,Song Lin Guo,et al.Feature extraction method of bearing AE signal based on improved FAST-ICAand wavelet packet energy Original[J].Mechanical Systems and Signal Processing,2015.62-63:91-99.
[2]Zhiyang Wang,Jin Chen,Guangming Dong,et al.Constrained independent component analysis and its application to machine fault diagnosis[J].Mechanical Systems and Signal Processing,2011.25(7):2501-2512.
[3]Lin,J.,A.Zhang,Fault feature separation using wavelet-ICA filter[J].NDT&E International,2005.38(6):421-427.Processing-Letters and Reviews,2005.6(1):1-57.
[4]Hyv?rinen,A.One-Unit Contrast Functions for Independent Component Analysis:A Statistical Analysis[J].Neural Networks for Signal Processing,1997.
[5]Lu,W.,J.C.Rajapakse,ICA with Reference.Neurocomputing[J],2006.69(16-18):2244-2257.
[66]Li,C.,G.Liao,A reference-based blind source extraction algorithm[J].Neural Computing&Applications,2010.19(2):299-303.
[7]Randall,R.B.,J.Antoni,S.Chobsaard,The relationship between special correlation and envelope analysis in the diagnostics of bearing faults and other cyclostationary machine signals[J].Mechanical Systems and Signal Processing,2001.15(5):945-962.
[2]Zhiyang Wang,Jin Chen,Guangming Dong,et al.Constrained independent component analysis and its application to machine fault diagnosis[J].Mechanical Systems and Signal Processing,2011.25(7):2501-2512.
[3]Lin,J.,A.Zhang,Fault feature separation using wavelet-ICA filter[J].NDT&E International,2005.38(6):421-427.Processing-Letters and Reviews,2005.6(1):1-57.
[4]Hyv?rinen,A.One-Unit Contrast Functions for Independent Component Analysis:A Statistical Analysis[J].Neural Networks for Signal Processing,1997.
[5]Lu,W.,J.C.Rajapakse,ICA with Reference.Neurocomputing[J],2006.69(16-18):2244-2257.
[66]Li,C.,G.Liao,A reference-based blind source extraction algorithm[J].Neural Computing&Applications,2010.19(2):299-303.
[7]Randall,R.B.,J.Antoni,S.Chobsaard,The relationship between special correlation and envelope analysis in the diagnostics of bearing faults and other cyclostationary machine signals[J].Mechanical Systems and Signal Processing,2001.15(5):945-962.