Prediction of bearing performance degradation trend based on CAE and TCN

PDF(4060 KB)

Journal of Vibration and Shock ›› 2023, Vol. 42 ›› Issue (13) : 214-225.

LIU Yuanbo, CHEN Xiang, LIU Yu

Author information +

History +

Abstract

Aiming at the shortcomings of existing performance degradation prediction research, such as information loss when building health indicator, poor parallel computing performance and small receptive field when building prediction models, combined with the timing characteristics of performance degradation of monitored objects, a performance degradation prediction method based on Convolutional Auto-Encoder (CAE) and Temporal Convolutional Network (TCN) is proposed. Firstly, construct a high-dimensional feature set in the multi-domain of vibration signals, and use comprehensive evaluation indicators to preliminarily screen performance degradation indicators with good sensitivity and strong trend. Secondly, the Kernel Principal Component Analysis (KPCA) method is adopted to eliminate redundant information between multi-domain features and realize the construction of health indicator based on CAE network. On this basis, a performance degradation prediction model based on TCN is constructed, and direct multi-step prediction is used to achieve performance degradation trend prediction, and the effectiveness of the method is verified by using the bearing public data set. The results show that the feature set can be reduced from 14 dimensions to 4 dimensions by using KPCA, while 97.63% of the information of the original feature set is retained. Furthermore, the method of constructing health indicator based on CAE network is effective. The change process of the constructed health indicator truly reflects the performance degradation process of the bearing. Compared with the two commonly used health index construction methods, Auto-Encoding (AE) network and Gaussian Mixture Model (GMM), this method has obvious advantages. At the same time, the prediction model based on the TCN algorithm can accurately predict the performance degradation of the bearing, which has better performance and higher prediction accuracy than the prediction model based on the Long Short-Term Memory (LSTM) network and the Gated Recurrent Unit (GRU). The root mean square error and mean absolute error when the prediction step size is 3 are 0.0257 and 0.0187, respectively. This method has general significance and may be extended to predict the performance degradation trend of other mechanical equipment/components.

Key words

Degradation Prediction / Feature Extraction / Kernel Principal Component Analysis / Health Indicator / Temporal Convolutional Networks / Convolutional Auto-Encoder

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

LIU Yuanbo, CHEN Xiang, LIU Yu. Prediction of bearing performance degradation trend based on CAE and TCN[J]. Journal of Vibration and Shock, 2023, 42(13): 214-225

References

[1] Ahmad W, Khan S A, Kim J M. A hybrid prognostics technique for rolling element bearings using adaptive predictive models[J]. IEEE Transactions on Industrial Electronics, 2017, 65(2): 1577-1584.
[2] Wu J, Wu C, Cao S, et al. Degradation data-driven time-to-failure prognostics approach for rolling element bearings in electrical machines[J]. IEEE Transactions on Industrial Electronics, 2018, 66(1): 529-539.
[3] 朱炬锟. 基于深度学习的滚动轴承性能退化趋势预测方法研究[D].重庆交通大学,2021.
ZHU Ju-kun. Research on Prediction Method of Rolling Bearing Performance Degradation Trend Based on Deep Learning [D]. Chongqing Jiaotong University, 2021.
[4] 刘俊华. 风电机组轴承性能退化趋势预测研究[D].华北电力大学,2020.
LIU Jun-hua. Research on Prediction of Performance Degradation Trend of Wind Turbine Bearings [D]. North China Electric Power University,2020.
[5] 王奉涛, 陈旭涛, 柳晨曦,等.基于KPCA和WPHM的滚动轴承可靠性评估与寿命预测[J].振动.测试与诊断,2017,37(03):476-483+626.
WANG Feng-tao, CHEN Xu-tao, LIU Chen-xi, et al. Reliability evaluation and life prediction of rolling bearing based on KPCA and WPHM[J]. Journal of Vibration , Measurement & Diagnosis, 2017,37(03): 476-483+626.
[6] 张雨琦, 邹金慧, 马军.多退化变量灰色预测模型的滚动轴承剩余寿命预测[J].探测与控制学报,2019,41(03):112-120.
ZHANG Yu-qi, ZOU Jin-hui, MA Jun. Rolling Bearing Residual Life Prediction Based on Grey Prediction Model with Multiple Degenerate Variables[J]. Journal of Detection & Control, 2019,41(03): 112-120.
[7] Qiao W, Ni X, Wang L, et al. New degradation feature extraction method of planetary gearbox based on alpha stable distribution[J]. Journal of Mechanical Science and Technology, 2021, 35(1): 1-19.
[8] Zhang B, Zhang S, Li W. Bearing performance degradation assessment using long short-term memory recurrent network [J]. Computers in Industry, 2019, 106: 14-29.
[9] 郑小霞,钱轶群,王帅.基于优选小波包与马氏距离的滚动轴承性能退化GRU预测[J].振动与冲击,2020,39(17):39-46+63.
ZHENG Xiao-xia, QIAN Yi-qun, WANG Shuai. GRU prediction for performance degradation of rolling bearings based on optimal wavelet packet and Mahalanobis distance[J]. Journal of Vibration and Shock, 2019, 106:14-29.
[10] 方博. 滚动轴承健康状态评估方法研究[D].哈尔滨工业大学,2019.
FANG Bo. Research on Health Status Assessment Method of Rolling Bearings[D]. Harbin Institute of Technology, 2019.
[11] 杨超,杨晓霞,李灵飞.基于灰色关联度和ELM的轴承性能退化趋势预测[J].组合机床与自动化加工技术,2019(11):105-108.
YANG Chao, YANG Xiao-xia, LI Ling-fei. Prediction of Bearing Performance Degradation Trend Based on Grey Relational Degree and ELM [J]. Modular Machine Tool & Automatic Manufacturing Technique, 2019(11): 105-108.
[12] 翟文睿. 采煤机性能退化评估研究[D].中国矿业大学,2020.
ZHAI Wen-rui. Study on Performance Degradation Evaluation of Shearer [D]. China University of Mining, 2020.
[13] 金棋,王友仁,王俊.基于深度学习多样性特征提取与信息融合的行星齿轮箱故障诊断方法[J].中国机械工程,2019,30(02):196-204.
JIN Qi, WANG You-ren, WANG Jun. Planetary Gearbox Fault Diagnosis Based on Multiple Feature Extraction and Information Fusion Combine with Deep Learning[J]. China Mechanical Engineering, 2019,30(02): 196-204.
[14] 陈祥龙,吴春志.基于1-DCNN的滚动轴承退化预测研究[J].兵器装备工程学报,2021,42(11):222-227.
CHEN Xiang-long, WU Chun-zhi. Prediction for Rolling Bearing Performance Degradation Based on 1-DCNN [J]. Chinese Journal of Ordnance Equipment Engineering, 2021, 42(11): 222-227.
[15] AlThobiani F, Ball A. An approach to fault diagnosis of reciprocating compressor valves using Teager–Kaiser energy operator and deep belief networks [J]. Expert Systems with Applications, 2014, 41(9): 4113-4122.
[16] 赵光权,葛强强,刘小勇,等.基于DBN的故障特征提取及诊断方法研究[J].仪器仪表学报,2016,37(09):1946-1953.
ZHAO Guang-quan, GE Qiang-qiang, LIU Xiao-yong, et al. Fault feature extraction and diagnosis method based on deep belief network[J]. Chinese Journal of Scientific Instrument, 2016, 37(09): 1946-1953.
[17] 戴邵武,陈强强,丁宇.基于时域特征的滚动轴承寿命预测[J].计算机测量与控制,2019,27(10):60-63.
DAI Shao-wu, CHEN Qiang-qiang, DING Yu. Prediction for Rolling Bearing Remaining Life Based on Time Domain Features[J]. Computer Measurement & Control, 2019, 27(10): 60-63.
[18] 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(02):183-189.
SHEN Zhong-jie, CHEN Xue-feng, HE Zheng-jia, et al. Remaining Life Predictions of Rolling Bearing Based on Relative Features and Multivariable Support Vector Machine[J]. Journal of Mechanical Engineering, 2013, 49(02): 183-189.
[19] 陈强强,戴邵武,戴洪德,等.基于SPA-FIG与优化ELM的滚动轴承性能退化趋势预测[J].振动与冲击,2020,39(19):187-194.
CHEN Qiang-qiang, DAI Shao-wu, DAI Hong-de, et al. Performance degradation trend prediction of rolling bearings based on SPA-FIG and optimized ELM [J]. Journal of Vibration and Shock, 2020, 39(19): 187-194.
[20] 梅文娟,高媛,杜立,等.基于在线相关熵极限学习机的器件退化趋势实时流预测方法[J].仪器仪表学报,2019,40(11):212-224.
MEI Wen-juan, GAO Yuan, DU Li, et al. Online sequential regularized correntropy criterion extreme learning machine on spark streaming signal prediction for electronic device degradation [J]. Chinese Journal of Scientific Instrument, 2019, 40(11): 212-224.
[21] 周福娜,高育林,王佳瑜,等.基于深度学习的缓变故障早期诊断及寿命预测[J].山东大学学报(工学版),2017,47(05):30-37.
ZHOU Fu-na, GAO Yu-lin, WANG Jia-yu, et al. Early Diagnosis and Life Prognosis for Slowly Varying Fault Based on Deep Learning [J]. Journal of Shan Dong University (Engineering Science), 2017, 47(05): 30-37.
[22] 张娜. 基于深度信念网络的滚动轴承寿命预测方法研究[D].湖南大学,2018.
ZHANG Na. A Method of Rolling Bearings Life Prediction Based on Deep Belief Network [D]. Hunan University, 2018.
[23] 杨宇,张娜,程军圣.全参数动态学习深度信念网络在滚动轴承寿命预测中的应用[J].振动与冲击,2019,38(10):199-205+249.
YANG Yu, ZHANG Na, CHENG Jun-sheng. Global parameters dynamic learning deep belief networks and its application in rolling bearing life prediction [J]. Journal of Vibration and Shock, 2019, 38(10): 199-205+249.
[24] Janssens O, Slavkovikj V, Vervisch B, et al. Convolutional neural network based fault detection for rotating machinery[J]. Journal of Sound and Vibration, 2016, 377: 331-345.
[25] 孙淑娥,姚柳,赵怡.基于卷积神经网络的时间序列数据融合算法[J].西安石油大学学报(自然科学版),2021,36(05):136-142.
SUN Shu-e, YAO Liu, ZHAO Yi. Time Series Data Fusion Algorithm Based on Convolutional Neural Network[J]. Journal of Xi'an Shiyou University(Natural Science Edition), 2021, 36(05): 136-142.
[26] 王鑫,吴际,刘超,等.基于LSTM循环神经网络的故障时间序列预测[J].北京航空航天大学学报,2018,44(04):772-784.
WANG Xin, WU Ji, LIU Chao, et al. Fault Time Series Prediction Based on LSTM Recurrent Neural Network [J]. Journal of Beijing University of Aeronautics and Astronautics, 2018, 44(04): 772-784.
[27] 李锋,陈勇,向往,等.基于量子加权长短时记忆神经网络的状态退化趋势预测[J].仪器仪表学报,2018,39(07):217-225.
LI Feng, CHEN Yong, XIANG Wang, et al. State degradation trend prediction based on quantum weighted long short-term memory neural network [J]. Chinese Journal of Scientific Instrument, 2018, 39(07): 217-225.
[28] 王鹏,邓蕾,汤宝平,等.基于自编码器和门限循环单元神经网络的滚动轴承退化趋势预测[J].振动与冲击,2020,39(17):106-111+133.
WANG Peng, DENG Lei, TANG Bao-ping, et al. Degradation trend prediction of rolling bearing based on auto-encoder and GRU neural network [J]. Journal of Vibration and Shock, 2020, 39(17): 106-111+133.
[29] 李江. 基于流形空间主曲线相似度的状态评估方法研究[D].重庆大学,2017.
LI Jiang. Research on State Evaluation Based on Similarity Analysis of Principal Curves in Manifold Space [D]. Chongqing University, 2017.
[30] Lei Y, Li N, Guo L, et al. Machinery health prognostics: A systematic review from data acquisition to RUL prediction[J]. Mechanical systems and signal processing, 2018, 104: 799-834.
[31] Schölkopf B, Smola A, Müller K R. Kernel principal component analysis[C]//International conference on artificial neural networks. Springer, Berlin, Heidelberg, 1997: 583-588.
[32] Lee J M, Yoo C K, Choi S W, et al. Nonlinear process monitoring using kernel principal component analysis[J]. Chemical engineering science, 2004, 59(1): 223-234.
[33] 于生宝,何建龙,王睿家,等.基于小波包分析和概率神经网络的电磁法三电平变换器故障诊断方法[J].电工技术学报,2016,31(17):102-112.
YU Sheng-bao, HE Jian-long,WANG Rui-jia, et al. Fault Diagnosis of Electromagnetic Three-level Inverter Based on Wavelet Packet Analysis and Probabiliatic Neural Networks [J].Transactions of China Electrotechnical Society, 2016,31(17):102-112.
[34] Bai S, Kolter J Z, Koltun V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling [J]. arXiv preprint arXiv:1803.01271, 2018.
[35] 于重重,宁亚倩,秦勇,等.基于T-SNE样本熵和TCN的滚动轴承状态退化趋势预测[J].仪器仪表学报,2019,40(08):39-46.
YU Chong-chong, NING Ya-qian, QIN Yong, et al. Prediction of rolling bearing state degradation trend based on T-SNE sample entropy and TCN [J]. Chinese Journal of Scientific Instrument, 2019,40(08):39-46.
[36] 刘致远,刘渊博,杨峰,等.Spark平台下基于优化BP神经网络模型的接触疲劳剩余寿命预测[J].重庆理工大学学报(自然科学),2021,35(10):111-119.
LIU Zhi-yuan, LIU Yuan-bo ,YANG Feng, et al. Remaining life prediction of contact fatigue based on optimized BP neural network model on Spark platform [J].Journal of Chongqing University of Technology(Natural Science),2021,35(10):111-119.