基于主成分分析和BP神经网络的钢丝绳断丝定量检测方法

摘要
图/表
参考文献(22)
相关文章 (15)

全文: PDF (841 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对目前钢丝绳断丝定量检测中存在的问题，充分利用主成分分析与BP神经网络的优点，提出了基于主成分分析与BP神经网络相结合的钢丝绳断丝定量检测方法。采用主成分分析法对钢丝绳断丝信号的原始特征属性进行预处理，得到钢丝绳断丝信号主成分特征属性，并以此作为BP神经网络的输入，建立钢丝绳断丝信号主成分特征属性与断丝数目之间的关系，并对钢丝绳断丝数目进行预测。主成分分析方法减少了原始特征属性的维数，消除了属性之间的相关性。同时，主成分特征属性作为BP神经网络的输入，也简化了网络的结构。实例测试结果表明，基于主成分分析的神经网络钢丝绳断丝检测方法与常规BP神经网络方法相比，具有更高的检测精度和更少的计算量。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章

关键词 ：钢丝绳, 断丝, 主成分分析, 神经网络

Abstract：In terms of the problems existing in steel rope broken wire quantitative testing, making full use of the advantages of principal component analysis (PCA) and BP neural network, a quantitative testing method for broken wire in steel rope was established.Inducting principal component analysis method to pre-analyze the original character attributes of broken wire signal for steel rope, and using the principal components of original character attributes as the input of BP neural network, the relationship between the principal components of original character attributes of broken wire signal for steel rope and the number of broken wire were estimated and the number of broken wire were predicted.Principal components method cut down the dimensions of original character attributes and eliminated the correlation among original character attributes.Meanwhile, the structure of BP neural network was also simplified by the principal components of original character attributes.Testing results show that BP neural network based on principal component analysis method compared with conventional BP neural network improves the prediction precision of broken wire in steel rope and reduces neural network training time.

Key words： steal rope broken wire principal component analysis artificial neural network model

收稿日期: 2017-05-18 出版日期: 2018-09-15

引用本文:

刘志怀1,2，秦芳3，刘娜4，黄祖坤1，刘学斌5. 基于主成分分析和BP神经网络的钢丝绳断丝定量检测方法[J]. 振动与冲击, 2018, 37(18): 271-276.
LIU Zhihuai1,2 QIN Fang3 LIU Na4 HUANG Zukun1 LIU Xubin5. Quantitative testing method for broken wire in steel rope based on principal component analysis and BP artificial neural network model. JOURNAL OF VIBRATION AND SHOCK, 2018, 37(18): 271-276.

链接本文:

http://jvs.sjtu.edu.cn/CN/ 或 http://jvs.sjtu.edu.cn/CN/Y2018/V37/I18/271

[1] 曹青松，刘丹，周继惠，等. 一种钢丝绳断丝无损定量检测方法[J]. 仪器仪表学报，2011，32(4)：787-794.
Cao Qingsong, Liu Dan, Zhou Jihui,et al. Non-destructive and quantitative detection method for broken wire rope [J]. Chinese Journal of Scientific Instrument, 2011, 32(4): 787-794.
[2] 张东来，徐殿国，王炎. B-小波的特点及其在钢丝绳断丝信号处理中的应用[J]. 哈尔滨工业大学学报，1998，30(4)：107-112.
Zhang Donglai, Xu Dianguo, Wang Yan. Feature of B-wavelet analysis and application in broken wires signal processing based on linear B-wavelet [J]. Journal of Harbin Institute of Technology, 1998, 30(4): 107-112.
[3] 江四厚，王汉功，阳能军，等. 小波分析在钢丝绳检测信号处理中的应用[J]. NDT无损检测，2006，28(2)：70-93.
Jiang Sihou, Wang Hangong, Yang Nengjun,et al. Application of wavelet analysis to the signal processing in wire rope testing [J]. Nondestructive Testing, 2006, 28(2): 70-93.
[4] 王阳生，师汉民，杨叔子. 钢丝绳断丝定量检测的原理与实现[J]. 中国科学(A辑)，1989，9：993-1000.
[5] 战卫侠. 钢丝绳断丝损伤信号处理及定量识别方法研究[D].青岛: 青岛理工大学，2013:5-8.
Zhan Weixia. Research on signal process and quantitative recognition method of broken wires in wire rope [D]. Qindao: Qingdao Technological University, 2013: 5-8.
[6] 陈厚桂，康宜华，武新军. 基于相关性的钢丝绳断丝特征提取[J]. 机械工程学报，2006，42(5)：224-228.
Chen Hougui, Kang Yihua, Wu Xinjun. Correlation analysis of broken wire signal in wire rope testing [J]. Chinese Journal of Mechanical Engineering, 2006, 42(5): 224-228.
[7] 胡阳，康宜华，卢文祥，等. 钢丝绳无损检测中的一些算法-信号的预处理和特征提取[J]. 无损检测，2000，22(11)：483-488.
Hu Yang, Kang Yihua, Lu Wenxiang,et al. Some algorithms for nodestructive testing of wire ropes-signal pre-processing and character extraction [J]. Nondestructive Testing , 2000, 22(11): 483-488.
[8] 徐俊峰，陶德馨. 基于BP神经网络的钢丝绳断丝模式识别[J].武汉理工大学学报，2002，24(2)：52-55.
Xu Junfeng, Tao Dexin. Module identification for broken wires of steel wire rope based on BP neural network [J]. Journal of Wuhan University of Technology, 2002, 24(2): 52-55.
[9] 谭继文. 钢丝绳损伤与张力在线定量检测主安全性评价的研究[D]. 沈阳: 东北大学，2000:76-82.
Tan Jiwen. Study on online quantitive test of wire-rope’s damage and tension and safety evaluation [D]. Shenyang: Northeastern University, 2000: 76-82.
[10] 张东来，徐殿国，王炎. 钢丝绳断丝信号的空间域划分及在小波变换下的频域特征提取[J]. 电子学报，2000，28(7)：59-62.
Zhang Donglai, Xu Dianguo, Wang Yan. Demarcation in space domain for local flaw signals of wire ropes and feature extraction in frequency domain based on wavelet transform [J]. Acta Electronica Sinica, 2000, 28(7): 59-62.
[11] 王军，谭继文, 战卫侠，等. 基于小波能量的钢丝绳断丝损伤信号处理[J]. 青岛理工大学学报，2012，33(3)：65-69.
Wang Jun, Tan Jiwen, Zhan Weixia,et al. Signal processing of broken wire damage in steel ropes based on wavelet energy [J]. Journal of Qindao Technological University, 2012, 33(3): 65-69.
[12] 张静远，张冰，蒋兴舟. 基于小波变换的特征提取方法分析[J]. 信号处理，2000，16(2)：156-162.
Zhang Jingyuan, Zhang Bing, Jiang Xingzhou. Analysis of feature extraction methods based on wavelet transform [J]. Signal Processing, 2000, 16(2): 156-162.
[13] Christen R, Bergamini A. Automatic flaw detection in NDE signals using a panel of neural network [J]. NDT&E International, 2006, 39: 547-553.
[14] 田志勇，张耀，谭继文. 基于BP神经网络的钢丝绳断丝定量检测[J]. 煤炭学报，2006，31(2)：245-249.
Tian Zhiyong, Zhang Yao, Tan Jiwen. Quantative test of broken wire for steel rope based on the back-propagation artificial neural networks [J]. Journal of China Coal Society, 2006, 31(2): 245-249.
[15] 程羽，张晓光，袁进南，等. 基于BP神经网络的钢丝绳断丝模式识别[J]. 煤矿安全，2009，2：88-91.
[16] 谭继文. 钢丝绳LF型断丝定量识别的神经网络法[J]. 矿山机械，2002，3：37-39.
Tan Jiwen. Nervous network method identifying cracked wires of a rope quantitatively [J]. Mining & Processing Equipment, 2002, 3: 37-39.
[17] 李春华，王璐. 小波分析与神经网络在钢丝绳断丝处理中的应用[J]. 自动化仪表，2009，30(12)：61-64.
Li Chunhua, Wang Lu. Application of wavelet analysis and neural network under handling wire rope broken condition [J]. Process Automation Instrumentation, 2009, 30(12): 61-64.
[18] Tang Wanmei. The study of the optimal structure of BP neural network [J]. System Engineering Theory and Practice, 2005, 25(10): 95-100.
[19] 贾磊，万百五，冯祖仁. 以高维输入神经网络作为生产线产品质量模型[J]. 控制与决策，2000，15(5)：569-572.
Jia Lei, Wan Baiwu, Feng Zuren. Neural network quality model with high-dimension inputs for production line products [J]. Control and Decision, 2000, 15(5): 569-572.
[20] 徐俊峰. 钢丝绳断丝损伤的智能化检测技术研究[D]. 武汉: 武汉理工大学，2002:44-50.
Xu Junfeng. The intelligent detection technique study for broken wires of the steel wire rope [D]. Wuhan: Wuhan University of Technology, 2002: 44-50.
[21] 周强，陶德馨，金超球. 钢丝绳断丝漏磁信号特征分析[J]. 港口装卸，2004，2：6-8.
[22] 王长龙，陈鹏，刘美全，等. 漏磁信号特征提取及检测研究[J]. 军械工程学院学报，2004，16(4)：1-4.
Wang Changlong, Chen Peng, Liu Meiquan,et al. Study on feature extraction and detection to magnetic flux leakage signals [J]. Journal of Ordnance Engineering College , 2004, 16(4): 1-4.