复杂装备轴承多重故障的线性判别分析与反向传播神经网络协作诊断方法

doi:10.3969/j.issn.1000-1093.2017.08.024

兵工学报 ›› 2017, Vol. 38 ›› Issue (8): 1649-1657.doi: 10.3969/j.issn.1000-1093.2017.08.024

复杂装备轴承多重故障的线性判别分析与反向传播神经网络协作诊断方法

黄大荣^1，2, 陈长沙¹，孙国玺²，赵玲¹，米波¹

(1.重庆交通大学信息科学与工程学院, 重庆 400074；2.广东石油化工学院广东省石化装备故障诊断重点实验室，广东茂名 525000)

收稿日期:2016-12-02 修回日期:2016-12-02 上线日期:2017-10-10
作者简介:黄大荣（1978—），男，教授，硕士生导师。E-mail：hcx1978@163.com
基金资助:
国家自然科学基金项目（61663008、61573076、61473094、61304104、61004118）；教育部留学归国人员科研启动基金项目（2015-49）;重庆市高等学校优秀人才支持计划项目（2014-18）；广东省石化装备故障诊断重点实验室开放式基金项目（GDUPKLAB201501、GDUPKLAB201604）；重庆市研究生教育教学改革研究重点项目（yjg152011）；重庆市高等教育学会2015—2016高等教育科学研究课题项目（CQGJ15010C）

Linear Discriminant Analysis and Back Propagation Neural Network Cooperative Diagnosis Method for Multiple Faults of ComplexEquipment Bearings

HUANG Da-rong^1,2, CHEN Chang-sha¹, SUN Guo-xi², ZHAO Ling¹，MI Bo¹

(1.College of Information Science and Engineering, Chongqing Jiaotong University, Chongqing 400074, China;2.Guangdong Provincial Key Laboratory of Petrochemical Equipment Fault Diagnosis, Guangdong University of Petrochemical Technology, Maoming 525000, Guangdong , China)

Received:2016-12-02 Revised:2016-12-02 Online:2017-10-10

摘要/Abstract

摘要： 由于复杂装备运行工作环境恶劣，导致其轴承多重故障诊断的准确率不高，为此提出一种基于线性判别分析(LDA)与反向传播(BP)神经网络协作下复杂装备轴承数据驱动的多重故障诊断方法。将无量纲指标作为轴承多重故障数据的反映指标，利用LDA对轴承多重故障的无量纲指标数据进行线性映射降维处理；通过拉格朗日极值法获得最佳投影向量，沿着该方向将轴承多重故障数据投影到类别最易区分的方向；将经投影处理后的样本作为BP神经网络的输入样本，通过训练测试网络，实现轴承多重故障的预测分类。对某型装备大型旋转机械机组进行仿真实验，验证了所提方法能够有效对轴承多重故障进行降维映射，并且能较好地实现多重故障分类诊断，具有良好的有效性和实用性。

关键词: 机械学, 轴承多重故障诊断, 拉格朗日极值法, 线性判别分析, 反向传播神经网络

Abstract: The fault diagnosis accuracy of bearing for complex equipment is not high due to the structural complexity of complex equipment and the poor working environment. A method of multiple bearing fault diagnosis based on linear discriminant analysis (LDA) and BP neural network is presented. A linear discriminant analysis is utilized for the linear dimension reduction of the dimensionless bearing multiple fault index, which is taken as an indicator of fault data. Lagrange extremum method is used to obtain an optimal projection vector. The bearing multiple fault data is projected on a category most likely to distinguished direction. The projected samples are used as the input samples of BP neural network and the test network. The simulation experiment of a certain large rotating machinery units shows that the proposed method can effectively reduce the dimensional mapping of multi-fault, achieve better classification, and has good validity and practicability. Key

Key words: mechanics, bearingmultiplefaultdiagnosis, Lagrangianextremummethod, lineardiscriminantanalysis, BPneuralnetwork

中图分类号:

TH133.33⁺1

黄大荣, 陈长沙，孙国玺，赵玲，米波. 复杂装备轴承多重故障的线性判别分析与反向传播神经网络协作诊断方法[J]. 兵工学报, 2017, 38(8): 1649-1657.

HUANG Da-rong, CHEN Chang-sha, SUN Guo-xi, ZHAO Ling，MI Bo. Linear Discriminant Analysis and Back Propagation Neural Network Cooperative Diagnosis Method for Multiple Faults of ComplexEquipment Bearings[J]. Acta Armamentarii, 2017, 38(8): 1649-1657.

参考文献

［1］崔玲丽, 张宇, 巩向阳, 等. 基于振动响应机理的轴承故障定量诊断及量化分析[J]. 北京工业大学学报, 2015, 41(11):1681-1687.
CUI Ling-li, ZHANG Yu, GONG Xiang-yang, et al. Vibration mechanism based quantitative diagnosis and quantization analysis of rolling bearing fault[J]. Journal of Beijing University of Technology, 2015, 41(11):1681-1687. (in Chinese)
[2］石双平. 滚动轴承故障诊断机理研究[J].机械设计与制造, 2014(10):236-239.
SHI Shuang-ping. Mechanism of rolling bearings fault diagnosis[J]. Machinery Design & Manufacture, 2014(10):236-239. (in Chinese)
[3］ Saimurugan M, Ramachandran K I, Sugumaran V. Multi component fault diagnosis of rotational mechanical system based on decision tree and support vector machine[J]. Expert Systems with Application, 2011, 38(4):3819-3826.
[4］ Prasad G D, Kumar B R, Ramana K V. Development of an expert system for diagnosis of bearing faults of rotating machinery with a case study on boiler feed pump[J]. International Journal of Current Research & Review, 2012, 4(7):16-18.
[5］ Li B, Chow M Y, Tipsuwan Y, et al. Neural-network-based motor rolling bearing fault diagnosis[J]. IEEE Transactions on Industrial Electronics, 2000, 47(5):1060-1069.
[6］王新峰, 邱静, 刘冠军. 基于绕封模型的故障特征选择方法研究[J]. 兵工学报, 2005, 26(5):685-689.
WANG Xin-feng, QIU Jing, LIU Guan-jun. Research on fault feature selection methods based on the sealing model[J]. Acta Armamentarii, 2005, 26(5):685-689.(in Chinese)
[7］张锐戈, 谭永红. 双谱主成分分析的滚动轴承智能故障诊断[J]. 振动工程学报, 2014, 27(5) :763-769.
ZHANG Rui-ge, TAN Yong-hong. Intelligent fault diagnosis of rolling element bearing based on bispectrum principal components analysis[J]. Journal of Vibration Engineering, 2014, 27(5):763-769. (in Chinese)
[8］程静, 王维庆, 何山. 双谱分析法在风机轴承故障诊断中的应用[J]. 自动化仪表, 2016, 37(7):27-29.
CHENG Jing, WANG Wei-qing, HE Shan. Application of the dual spectrum analysis method in fault diagnosis of wind turbine bearing[J]. Process Automation Instrumentation, 2016, 37(7):27-29.(in Chinese)
[9］郝腾飞, 陈果, 廖仲坤. 基于正则化多核判别分析的航空发动机滚动轴承早期故障融合诊断方法[J]. 航空动力学报, 2013, 28(12):2759-2770.
HAO Teng-fei, CHEN Guo, LIAO Zhong-kun. Approach for incipient fusion fault diagnosis of rolling bearing of aero-engine based on regularized multiple kernel discriminant analysis[J]. Journal of Aerospace Power, 2013, 28(12):2759-2770.(in Chinese)
[10］杨杰, 郑海起, 关贞珍. 基于形态成分分析的轴承复合故障诊断[J]. 轴承, 2011(8):38-42.
YANG Jie, ZHENG Hai-qi, GUAN Zhen-zhen. Compound fault diagnosis for bearings based on morphological component analysis[J]. Bearing, 2011(8):38-42.(in Chinese)
[11］杨宇, 于德介, 程军圣. 基于Hilbert边际谱滚动轴承故障诊断方法[J]. 振动与冲击, 2005, 24(1):70-72.
YANG Yu, YU De-jie, CHENG Jun-sheng. Roller bearing fault diagnosis based on Hilbert marginal spectrum[J]. Journal of Vibration and Shock, 2005, 24(1):70-72.(in Chinese)
[12］梁峰, 张丹丹, 张荣华. 基于BP神经网络的水润滑艉轴承故障诊断研究[J]. 机械传动, 2015, 39(7):118-121.
LIANG Feng, ZHANG Dan-dan, ZHANG Rong-hua. Fault diagnosis of water lubricated stern bearing based on BP neural network[J]. Journal of Mechanical Transmission, 2015, 39(7):118-121.(in Chinese)
[13］赵刚, 黄大荣, 黄席樾. 火箭漏电故障诊断技术研究[J].兵工学报, 2010, 31(7):916-921.
ZHAO Gang, HUANG Da-rong, HUANG Xi-yue. Research on the leak current fault diagnosis technology for launch vehicle[J]. Acta Armamentarii, 2010, 31(7):916-921.(in Chinese)
[14］梁传建, 杨国来, 王晓锋. 基于神经网络和遗传算法的火炮结构动力学优化[J]. 兵工学报, 2015, 36(5):789-794.
LIANG Chuan-jian, YANG Guo-lai, WANG Xiao-feng. Structural dynamics optimization of gun based on neural networks and genetic algorithms[J]. Acta Armamentarii, 2015, 36(5):789-794.(in Chinese)
[15］张可, 柴毅, 刘建环. 一类基于类别区分的多重故障诊断模型框架[J]. 控制理论与应用, 2016, 33(2):154-163.
ZHANG Ke, CHAI Yi, LIU Jian-huan. A class of category discrimination based model frameworks for multiple fault diagnosis[J]. Control Theory & Applications, 2016, 33(2):154-163.(in Chinese)
[16］张清华, 邵龙秋, 李红芳. 基于无量纲指标的旋转机械并发故障诊断技术[J]. 华中科技大学学报:自然科学版, 2009, 37(增刊1):156-159.
ZHANG Qing-hua,SHAO Long-qiu,LI Hong-fang.Research of concurrent fault diagnosis technology of rotating machinery based on the non-dimensional parameter[J]. Journal of Huazhong University of Science and Technology: Natural Science,2009, 37(S1): 156-159. (in Chinese)
[17］李志农, 朱明.基于变分模态分解的机械故障诊断方法研究[J].兵工学报,2017,38(3):593-599.
LI Zhi-nong,ZHU Ming.Research on mechanical fault diagnosis method based on variational mode decomposition[J]. Acta Armamentarii, 2017,38(3):593-599. (in Chinese)
[18］程昊宇, 董朝阳, 江未来,等.变体飞行器故障检测与容错控制一体化设计[J].兵工学报,2017,38(4):711-721.
CHENG Hao-yu,DONG Chao-yang,JIANG Wei-lai,et al.Integrated fault diagnosis and fault tolerant control for morphing aircraft[J].Acta Armamentarii,2017,38(4):711-721. (in Chinese)
[19］李海广，潘宏侠，任海锋.基于冲击响应谱特征提取的自动机裂纹故障诊断[J].兵工学报2016,37(9):1744-1752.
LI Hai-guang,PAN Hong-xia,REN Hai-feng.Crack fault diagnosis of automatic mechanism based on shock response spectrum features extraction[J]. Acta Armamentarii, 2016,37(9):1744- 1752. (in Chinese)
[20］田恒，段富海，江秀红，等.基于准信息熵的测试性D矩阵故障诊断新算法[J].兵工学报，2016,37(5):923-928.
TIAN Heng,DUAN Fu-hai,JIANG Xiu-hong,et al.A novel fault diagnosis algorithm of testability D matrix based on quasi information entropy[J]. Acta Armamentarii, 2016,37(5):923-928. (in Chinese)

第38卷
第8期2017 年8月兵工学报ACTA
ARMAMENTARIIVol.38No.8Aug.2017

[1]	刘冰, 郝新红, 周文, 杨瑾. 基于BAS-BPNN的调频无线电引信目标与扫频干扰识别方法[J]. 兵工学报, 2023, 44(8): 2391-2403.
[2]	李晓雄，张淑宁，赵惠昌，陈思. 基于1DC-CGAN和小波能量特征的引信小样本地形目标识别[J]. 兵工学报, 2022, 43(10): 2545-2553.
[3]	赵星宇，白春华，姚箭，孙彬峰. 燃料空气炸药爆轰产物JWL状态方程参数计算[J]. 兵工学报, 2020, 41(10): 1921-1929.
[4]	何耀民，何华锋，徐永壮，苏敬，王依繁. 基于反向传播神经网络的海杂波参数估计[J]. 兵工学报, 2019, 40(12): 2473-2481.
[5]	高玉波，张伟, 李达诚，宜晨虹, 汤铁钢. 基于反向传播神经网络的陶瓷损伤参数反演分析[J]. 兵工学报, 2018, 39(1): 146-152.
[6]	邱明，周大威，庞晓旭. 织物衬垫编织工艺对自润滑关节轴承摩擦学性能的影响[J]. 兵工学报, 2017, 38(9): 1867-1872.
[7]	曲梦可，王洪波，荣誉. 轮腿混合机器人机械腿动力学建模与驱动预估[J]. 兵工学报, 2017, 38(8): 1619-1629.
[8]	牛荣军，徐金超，邵秀华，邓四二. 双排非对称四点接触球转盘轴承刚度分析[J]. 兵工学报, 2017, 38(6): 1239-1248.
[9]	谢宏伟，陶忠，侯军占，张卫国. 多绳精密传动的传动精度实验研究[J]. 兵工学报, 2017, 38(4): 728-734.
[10]	胡俊峰，蔡建阳. 基于振动控制的微小物体释放操作方法与实验研究[J]. 兵工学报, 2017, 38(4): 802-809.
[11]	谢克峰，张合，刘善增，李豪杰. 海上柔性并联平台空间动力学建模与分析[J]. 兵工学报, 2017, 38(3): 512-519.
[12]	李志农，朱明. 基于变分模态分解的机械故障诊断方法研究[J]. 兵工学报, 2017, 38(3): 593-599.
[13]	李世尧，张卫国，侯军占，段文博，迟圣威，郭冰涛. 减振器动刚度特性研究[J]. 兵工学报, 2017, 38(11): 2274-2279.
[14]	邱明，梁霞, 李迎春. 注塑改性尼龙杆端关节轴承的摩擦学性能[J]. 兵工学报, 2017, 38(10): 2075-2080.
[15]	萧辉，杨国来，孙全兆. 火炮多柔体动力学结构优化研究[J]. 兵工学报, 2017, 38(1): 27-34.

复杂装备轴承多重故障的线性判别分析与反向传播神经网络协作诊断方法

Linear Discriminant Analysis and Back Propagation Neural Network Cooperative Diagnosis Method for Multiple Faults of ComplexEquipment Bearings

PDF

可视化

被引次数

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价