Bearing Fault Diagnosis based on Deep Belief Networks and Particle Swarm Optimization Support Vector Machine
-
摘要: 针对如何提高轴承故障诊断的准确率和算法训练的效率问题,提出了一种深度信念网络(DBN)与粒子群优化支持向量机(PSO-SVM)的滚动轴承故障诊断方法。首先,求出信号的时频特征统计量,其次,利用DBN对时频特征统计量进行特征提取,最后,利用PSO-SVM进行分类。实验结果表明:相比于直接用PSO-SVM进行分类,该方法不仅准确率更高,而且算法训练的时间大大缩短了,提高了滚动轴承故障诊断的准确率和效率。Abstract: As how to improve the accuracy and algorithm efficiency of roll bearing fault diagnosis, a new method of bearing fault diagnosis based on deep belief network (DBN) and the particle swarm optimization support vector machine (PSO-SVM) with the time-frequency characteristic statistic is proposed. Firstly, time-frequency characteristic statistic of the bearing vibration signal is calculated. And then the DBN is used to extract features of time-frequency feature extraction. Finally, the extracted parameters are input to the PSO-SVM to be classified. The experimental results show that this method not only has higher accuracy, but also greatly shorten the training time, and the accuracy and efficiency of fault diagnosis is improved as a result.
-
表 1 时域特征参量
特征参数 表达式 注意 平均值 有效值 xi是信号x的第i个值, N是数据总数 波形因子 峰峰值 Pk=max[x]-min[x] 波峰因子 峭度 σ2为方差 峭度因子 脉冲因子 I=PK/(av) av为绝对值的平均 Xr Xr=mean{sqrt[abs(X)]}2 裕度因子 L=PK/xr St 表 2 频域特征参量
特征参数 表达式 注意 频度重心 RMS变异频率 Root变异频率 表 3 轴承数据集
故障类型 训练/测试样本 数据集 类别编号 正常 100/100 97DE 000 内圈故障(0.177 8) 100/100 105DE 001 内圈故障(0.355 6) 100/100 169DE 010 内圈故障(0.533 4) 100/100 211DE 011 外圈故障(0.177 8) 100/100 130DE 100 外圈故障(0.355 6) 100/100 198DE 101 外圈故障(0.533 4) 100/100 236DE 110 表 4 分类器性能对照表
方法 运行时间/s (c, g) 准确率/% PSO-SVM 27.36 (7.093 1, 5.239 0) 84.85 DBN-PSO-SVM 6.62 (158.953 5, 31.818 3) 97.34 -
[1] Tran V T, Yang B S, Gu F S, et al. Thermal image enhancement using bi-dimensional empirical mode decomposition in combination with relevance vector machine for rotating machinery fault diagnosis[J]. Mechanical Systems and Signal Processing, 2013, 38(2):601-614 doi: 10.1016/j.ymssp.2013.02.001 [2] Shen C Q, Wang D, Kong F R, et al. Fault diagnosis of rotating machinery based on the statistical parameters of wavelet packet paving and a generic support vector regressive classifier[J]. Measurement, 2013, 46(4):1551-1564 doi: 10.1016/j.measurement.2012.12.011 [3] Hinton G E, Osindero S, Teh Y W. A fast learning algorithm for deep belief nets[J]. Neural Computation, 2006, 18(7):1527-1554 doi: 10.1162/neco.2006.18.7.1527 [4] 李巍华, 单外平, 曾雪琼.基于深度信念网络的轴承故障分类识别[J].振动工程学报, 2016, 29(2):340-347 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=zdgcxb201602020Li W H, Shan W P, Zeng X Q. Bearing fault identification based on deep belief network[J]. Journal of Vibration Engineering, 2016, 29(2):340-347(in Chinese) http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=zdgcxb201602020 [5] 张淑清, 胡永涛, 姜安琦, 等.基于双树复小波和深度信念网络的轴承故障诊断[J].中国机械工程, 2017, 28(5):532-536, 543 doi: 10.3969/j.issn.1004-132X.2017.05.005Zhang S Q, Hu Y T, Jiang A Q, et al. Bearing fault diagnosis based on DTCWT and DBN[J]. China Mechanical Engineering, 2017, 28(5):532-536, 543(in Chinese) doi: 10.3969/j.issn.1004-132X.2017.05.005 [6] 单外平, 曾雪琼.基于深度信念网络的信号重构与轴承故障识别[J].电子设计工程, 2016, 24(4):67-71 doi: 10.3969/j.issn.1674-6236.2016.04.023Shan W P, Zeng X Q. Signal reconstruction and bearing fault identification based on deep belief network[J]. Electronic Design Engineering, 2016, 24(4):67-71(in Chinese) doi: 10.3969/j.issn.1674-6236.2016.04.023 [7] 单外平.基于深度信念网络的变速器故障分类识别研究[D].广州: 华南理工大学, 2015 http://cdmd.cnki.com.cn/Article/CDMD-10561-1015987240.htmShan W P. Classification and recognition of transmission fault based on deep belief network[D]. Guangzhou: South China University of Technology, 2015(in Chinese) http://cdmd.cnki.com.cn/Article/CDMD-10561-1015987240.htm [8] 俞昆.基于深度信念网络与多传感器信息融合的滚动轴承故障诊断研究[D].山东青岛: 青岛理工大学, 2016 http://cdmd.cnki.com.cn/Article/CDMD-10429-1017019033.htmYu K. Fault diagnosis of rolling element bearings based on deep belief network and multiple sensors information fusion[D]. Shandong Qingdao: Qingdao University of Technology, 2016(in Chinese) http://cdmd.cnki.com.cn/Article/CDMD-10429-1017019033.htm [9] 雷亚国, 贾峰, 周昕, 等.基于深度学习理论的机械装备大数据健康监测方法[J].机械工程学报, 2015, 51(21):49-56 http://d.old.wanfangdata.com.cn/Periodical/jxgcxb201521007Lei Y G, Jia F, Zhou X, et al. A deep learning-based method for machinery health monitoring with big data[J]. Journal of Mechanical Engineering, 2015, 51(21):49-56(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jxgcxb201521007 [10] 李艳峰, 王新晴, 张梅军, 等.基于奇异值分解和深度信度网络多分类器的滚动轴承故障诊断方法[J].上海交通大学学报, 2015, 49(5):681-686, 694 http://d.old.wanfangdata.com.cn/Periodical/shjtdxxb201505018Li Y H, Wang X Q, Zhang M J, et al. An approach to fault diagnosis of rolling bearing using SVD and Multiple DBN classifiers[J]. Journal of Shanghai Jiaotong University, 2015, 49(5):681-686, 694(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/shjtdxxb201505018 [11] 陈园艺, 孙建平.基于EEMD和PSO-SVM的滚动轴承故障诊断[J].电力科学与工程, 2016, 32(10):47-52 doi: 10.3969/j.issn.1672-0792.2016.10.008Chen Y Y, Sun J P. Fault diagnosis of rolling bearing based on EEMD and PSO-SVM[J]. Electric Power Science and Engineering, 2016, 32(10):47-52(in Chinese) doi: 10.3969/j.issn.1672-0792.2016.10.008 [12] 李丹峰, 郭雄伟.基于双谱和PSO-SVM的滚动轴承故障诊断[J].贵州大学学报, 2011, 28(4):85-89 doi: 10.3969/j.issn.1000-5269.2011.04.022Li D F, Guo X W. Rolling bearing fault diagnosis based on bispectrum and PSO-SVM[J]. Journal of Guizhou University, 2011, 28(4):85-89(in Chinese) doi: 10.3969/j.issn.1000-5269.2011.04.022 [13] Chen F F, Tang B P, Song T, et al. Multi-fault diagnosis study on roller bearing based on multi-kernel support vector machine with chaotic particle swarm optimization[J]. Measurement, 2014, 47:576-590 doi: 10.1016/j.measurement.2013.08.021 [14] Sun W J, Shao S Y, Zhao R, et al. A sparse auto-encoder-based deep neural network approach for induction motor faults classification[J]. Measurement, 2016, 89:171-178 doi: 10.1016/j.measurement.2016.04.007 [15] 张小龙, 张氢, 秦仙蓉, 等.基于ITD复杂度和PSO-SVM的滚动轴承故障诊断[J].振动与冲击, 2016, 35(24):102-107, 138 http://d.old.wanfangdata.com.cn/Periodical/zdycj201624017Zhang X L, Zhang Q, Qin X R, et al. Rolling bearing fault diagnosis based on ITD Lempel-Ziv complexity and PSO-SVM[J]. Journal of Vibration and Shock, 2016, 35(24):102-107, 138(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/zdycj201624017 -