基于自联想神经网络与模糊C均值的滚动轴承的性能退化评估
|
摘要
滚动轴承是旋转机械中最重要也是最容易出现故障的零部件之一,如果能对滚动轴承的性能进行实时监测评估就能及时做出维修策略,故建立了自回归(AR)模型,提取滚动轴承全寿命周期的AR模型的自回归系数和残差,对提取到的特征降维后建立自联想神经网络(AANN)以及FCM模型,然后将AANN模型的输出与输入向量之差作为特征向量输入到FCM模型中,得到性能退化指标,再用实例对结论进行验证。实验表明,文中提出的性能退化方法得到的结论与轴承加速疲劳试验得到的结果是一致的。
Rolling bearing is one of the most important and easiest faulty components in rotating machinery. If we can monitor and estimate the performance of rolling bearing in real-time,we can make the maintenance strategy in time. The paper sets up autoregressive(AR) model and get the autoregressive coefficients and residualsof the life cycle of rolling bearing. The Auto-Associative Neural Networkmodel(AANN) and the Fuzzy C-meansmodel(FCM) are established by using the characteristics which was extracted and reduced to dimension. The feature vectorsof the differences between the output and input vectors of the AANN model are input into the FCM model. The performance degradation index is obtained. Then the instance is used to verify the conclusions of this paper. The experimental results show that the performance degradation method presented in this paper is consistent with the results obtained from the accelerated fatigue test.
Rolling bearing is one of the most important and easiest faulty components in rotating machinery. If we can monitor and estimate the performance of rolling bearing in real-time,we can make the maintenance strategy in time. The paper sets up autoregressive(AR) model and get the autoregressive coefficients and residualsof the life cycle of rolling bearing. The Auto-Associative Neural Networkmodel(AANN) and the Fuzzy C-meansmodel(FCM) are established by using the characteristics which was extracted and reduced to dimension. The feature vectorsof the differences between the output and input vectors of the AANN model are input into the FCM model. The performance degradation index is obtained. Then the instance is used to verify the conclusions of this paper. The experimental results show that the performance degradation method presented in this paper is consistent with the results obtained from the accelerated fatigue test.
引文
[1]周建民,郭慧娟,张龙.基于非线性降维和模糊均值聚类的滚动轴承的性能退化在线评估方法[J].机械设计与研究,2017,33(6):86-89.
[2]申中杰,陈雪峰,何正嘉,等.基于相对特征和多变量支持向量机的滚动轴承剩余寿命预测[J].机械工程学报,2013,49(2):183-189.
[3]LYBECK N S,MARBLE B.Morton.validating prognostic algorithms:A case study using comprehensive bearing fault data.in Aerospace Conference[J].IEEE,2007,Big Sky Montana:1-9.
[4]陈法法,杨勇,马婧华,等.信息熵与优化LS-SVM的轴承性能退化模糊粒化预测[J].仪器仪表学报,2016,37(4):779-787.
[5]何学文,孙林,付静.基于小波分析和支持向量机的旋转机械故障诊断方法[J],中国工程机械学报,2007,5(2):86-90
[6]LIU T,CHEN J,DONG G M.Zero crossing and coupled hidden Markov model for a rolling bearingperformance degradation assessment[J].Journal of Vibration and Control,2014,20(16):2487-2500.
[7]王奉涛,陈旭涛,闫达文,等.流形模糊C均值方法及其在滚动轴承性能退化评估中的应用[J].机械工程学报,2016,52(15):59-64.
[8]周建民,徐清瑶,张龙,等.基于小波包Tsallis熵和FCM的滚动轴承性能退化评估[J].机械传动,2016,40(5):110-115.
[9]姜海燕.基于小波与AR模型融合的滚动轴承故障诊断方法[J].工业控制计算机,2017,30(8):46-47
[10]曾小军,黄宜坚.基于AR模型和支持向量机的故障诊断法[J].机械科学与技术,2010,29(7):972-975.
[11]许跃,朱伟兴.基于自联想神经网络的传感器故障诊断[J].传感器与微系统,2007,26(4):94-96.
[12]司风琪,徐治皋.基于自联想神经网络的测量数据自校正检验方法[J].中国电机工程学报,2002,22(6):152-155.
[13]LEE J,QIU H,YU G,et al.Bearing data set[EB/OL].NASAAmes Prognostics Data Repository,2007[2015-05-04].http://ti.Arc.nasa.gov/tech/dash/pcoe/prognostic-data-repository.
[14]张龙,黄文艺,熊国良等.基于TESPAR与GMM的滚动轴承性能退化评估[J].仪器仪表学报,2014,35(8):1772-1779.
[15]ZHOU J M,GUO H J.Bearing performance degradation assessment using lifting wavelet packet symbolic entropy and SVDD[J].Shock and Vibration,2016,Article ID 3086454,10 pages,doi:10.1155/2016/3086454.
[2]申中杰,陈雪峰,何正嘉,等.基于相对特征和多变量支持向量机的滚动轴承剩余寿命预测[J].机械工程学报,2013,49(2):183-189.
[3]LYBECK N S,MARBLE B.Morton.validating prognostic algorithms:A case study using comprehensive bearing fault data.in Aerospace Conference[J].IEEE,2007,Big Sky Montana:1-9.
[4]陈法法,杨勇,马婧华,等.信息熵与优化LS-SVM的轴承性能退化模糊粒化预测[J].仪器仪表学报,2016,37(4):779-787.
[5]何学文,孙林,付静.基于小波分析和支持向量机的旋转机械故障诊断方法[J],中国工程机械学报,2007,5(2):86-90
[6]LIU T,CHEN J,DONG G M.Zero crossing and coupled hidden Markov model for a rolling bearingperformance degradation assessment[J].Journal of Vibration and Control,2014,20(16):2487-2500.
[7]王奉涛,陈旭涛,闫达文,等.流形模糊C均值方法及其在滚动轴承性能退化评估中的应用[J].机械工程学报,2016,52(15):59-64.
[8]周建民,徐清瑶,张龙,等.基于小波包Tsallis熵和FCM的滚动轴承性能退化评估[J].机械传动,2016,40(5):110-115.
[9]姜海燕.基于小波与AR模型融合的滚动轴承故障诊断方法[J].工业控制计算机,2017,30(8):46-47
[10]曾小军,黄宜坚.基于AR模型和支持向量机的故障诊断法[J].机械科学与技术,2010,29(7):972-975.
[11]许跃,朱伟兴.基于自联想神经网络的传感器故障诊断[J].传感器与微系统,2007,26(4):94-96.
[12]司风琪,徐治皋.基于自联想神经网络的测量数据自校正检验方法[J].中国电机工程学报,2002,22(6):152-155.
[13]LEE J,QIU H,YU G,et al.Bearing data set[EB/OL].NASAAmes Prognostics Data Repository,2007[2015-05-04].http://ti.Arc.nasa.gov/tech/dash/pcoe/prognostic-data-repository.
[14]张龙,黄文艺,熊国良等.基于TESPAR与GMM的滚动轴承性能退化评估[J].仪器仪表学报,2014,35(8):1772-1779.
[15]ZHOU J M,GUO H J.Bearing performance degradation assessment using lifting wavelet packet symbolic entropy and SVDD[J].Shock and Vibration,2016,Article ID 3086454,10 pages,doi:10.1155/2016/3086454.