基于相关系数和BP神经网络的轴承剩余寿命预测
|
摘要
为了评估机械设备的退化状态,准确掌握轴承剩余寿命信息,提出了一种基于相关系数和BP神经网络(BP Neural Network,BPNN)模型的轴承剩余寿命预测方法。该方法包括线上和线下两个步骤,首先利用相关系数法对预测模型的输入特征进行约简,其次线下步骤是通过机器学习来构建轴承的退化模型,而线上步骤则是利用退化模型来预测轴承剩余寿命。通过对轴承的全寿命退化实验数据进行分析预测,该方法与传统的技术相比能降低预测误差,表明该方法能够有效地模拟轴承退化过程并预测轴承剩余有效寿命(RUL)。
In order to evaluate the degradation state of the mechanical equipment and master the information of the remaining life of the bearing accurately, a method for predicting the remaining life of bearings based on correlation coefficient and BP neural network(BPNN)model is presented. The proposed method includes two steps of online and offline, first use the correlation coefficient to reduce features of the prediction model, then the offline step is used to build a degradation model of the bearing by machine learning and the online step predicts the remaining life by using the degeneration model. By analyzing the experimental data of bearing full lifetime degradation, the method can reduce the prediction error compared with traditional technology, the results show that the method can effectively simulate the bearing degradation process and predict the remaining useful life(RUL) of the bearing.
In order to evaluate the degradation state of the mechanical equipment and master the information of the remaining life of the bearing accurately, a method for predicting the remaining life of bearings based on correlation coefficient and BP neural network(BPNN)model is presented. The proposed method includes two steps of online and offline, first use the correlation coefficient to reduce features of the prediction model, then the offline step is used to build a degradation model of the bearing by machine learning and the online step predicts the remaining life by using the degeneration model. By analyzing the experimental data of bearing full lifetime degradation, the method can reduce the prediction error compared with traditional technology, the results show that the method can effectively simulate the bearing degradation process and predict the remaining useful life(RUL) of the bearing.
引文
[1] 张丹,隋文涛,黄雪梅.MED和WPT技术在轴承故障检测中的应用[J].组合机床与自动化加工技术,2016(9):51-53.
[2] 许立,张宇,汤武初,等.基于BP神经网络的圆锥滚子轴承故障诊断[J].组合机床与自动化加工技术,2016(11):105-111.
[3] 张丹,隋文涛,郭前建.采用Gabor滤波器的轴承状态监控方法[J].北京邮电大学学报,2016,39(2):103-107.
[4] T Benkedjouh,K Medjaher,N Zerhouni,et al.Remaining useful life estimation based on nonlinear feature reduction and support vector regression[J].Engineering Applications of Artificial Intelligence,2013,26:1751-1760.
[5] 康守强,叶立强,王玉静,等,Mikulovich V I.基于MCEA-KPCA和组合SVR的滚动轴承剩余使用寿命预测[J].电子测量与仪器学报,2017,31(9):1365-1371.
[6] 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.
[7] 雷亚国,吴雄辉,陈吴,等.一种基于相关向量机的滚动轴承运行趋势多步预测方法[C]//全国设备监测诊断与维护学术会议、全国设备故障诊断学术会议暨2014年全国设备诊断工程会议,2014.
[8] 裴峻峰,孙建华,宋传智,等.基于EEMD消噪和相关系数识别的滚动轴承故障诊断方法[J].机械传动,2018,42(4):150-155.
[9] 苏杭.基于大数据背景的相关系数[J].电子技术与软件工程,2018(7):182.
[10] 魏效玲,张宝刚,杨富贵,等.基于GA-BP网络的数控机床热误差优化建模研究[J].组合机床与自动化加工技术,2016(12):100-102.
[11] 龙泉,刘永前,杨勇平.基于粒子群优化BP神经网络的风电机组齿轮箱故障诊断方法[J].太阳能学报,2012,33(1):120-125.
[12] Nectoux P,Gouriveau,Medjaher K,et al.Pronostia:An experiment platform for bearings accelerated degradation tests[C].Proceedings of IEEE International Conference on Prognostics and Health Management,2012:1-8.
[2] 许立,张宇,汤武初,等.基于BP神经网络的圆锥滚子轴承故障诊断[J].组合机床与自动化加工技术,2016(11):105-111.
[3] 张丹,隋文涛,郭前建.采用Gabor滤波器的轴承状态监控方法[J].北京邮电大学学报,2016,39(2):103-107.
[4] T Benkedjouh,K Medjaher,N Zerhouni,et al.Remaining useful life estimation based on nonlinear feature reduction and support vector regression[J].Engineering Applications of Artificial Intelligence,2013,26:1751-1760.
[5] 康守强,叶立强,王玉静,等,Mikulovich V I.基于MCEA-KPCA和组合SVR的滚动轴承剩余使用寿命预测[J].电子测量与仪器学报,2017,31(9):1365-1371.
[6] 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.
[7] 雷亚国,吴雄辉,陈吴,等.一种基于相关向量机的滚动轴承运行趋势多步预测方法[C]//全国设备监测诊断与维护学术会议、全国设备故障诊断学术会议暨2014年全国设备诊断工程会议,2014.
[8] 裴峻峰,孙建华,宋传智,等.基于EEMD消噪和相关系数识别的滚动轴承故障诊断方法[J].机械传动,2018,42(4):150-155.
[9] 苏杭.基于大数据背景的相关系数[J].电子技术与软件工程,2018(7):182.
[10] 魏效玲,张宝刚,杨富贵,等.基于GA-BP网络的数控机床热误差优化建模研究[J].组合机床与自动化加工技术,2016(12):100-102.
[11] 龙泉,刘永前,杨勇平.基于粒子群优化BP神经网络的风电机组齿轮箱故障诊断方法[J].太阳能学报,2012,33(1):120-125.
[12] Nectoux P,Gouriveau,Medjaher K,et al.Pronostia:An experiment platform for bearings accelerated degradation tests[C].Proceedings of IEEE International Conference on Prognostics and Health Management,2012:1-8.