Residual life prediction method of rolling bearing based on DW-CNN-LSTM networks

PDF(2794 KB)

Journal of Vibration and Shock ›› 2022, Vol. 41 ›› Issue (17) : 282-291.

JIANG Quansheng1, XU Weiyang1, ZHU Junjun1, SHEN Yehu1, XU Fengyu2

Author information +

History +

Abstract

In the prediction of the remaining useful life of rolling bearings, data-driven methods cannot effectively extract the feature information sensitive to the degradation process of bearings, resulting in insufficient prediction accuracy. Therefore, a RUL prediction method of rolling bearing based on dynamic weighted convolution long short-term memory network (DW-CNN-LSTM) is proposed. Firstly, the rolling bearing vibration signal is decomposed by wavelet packet, and then the decomposed wavelet packet coefficient matrix is dynamically weighted by the dynamic weighting layer of trainable parameters to effectively screen the information that can characterize the bearing degradation, so as to enhance the feature learning ability of the model to the bearing vibration signal. Then, using the ability of convolutional neural network to adaptively learn and mine the deep features of data, the feature set sensitive to the bearing degradation process is extracted from the dynamically weighted wavelet packet coefficient matrix. Finally, with the advantage of long short-term memory (LSTM) to predict time information series, the high-dimensional degradation characteristics are further extracted through the double-layer LSTM to improve the prediction accuracy of rolling bearing RUL. The experimental results on XJTU-SY bearing dataset and IMS bearing dataset show that the proposed DW-CNN-LSTM method is better than the classical LSTM method in bearing life prediction accuracy, especially the root mean square error index of the proposed method is reduced by 61.08%, the prediction accuracy is improved by 9.95%, and the average training time of the model was reduced by 44.14%.
Key words：Rolling bearing; Prediction of the remaining useful life; Convolution long short-time memory network; Wavelet packet decomposition; Dynamic weighting

Key words

Rolling bearing / Prediction of the remaining useful life / Convolution long short-time memory network / Wavelet packet decomposition / Dynamic weighting

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

JIANG Quansheng1, XU Weiyang1, ZHU Junjun1, SHEN Yehu1, XU Fengyu2. Residual life prediction method of rolling bearing based on DW-CNN-LSTM networks[J]. Journal of Vibration and Shock, 2022, 41(17): 282-291

References

[1] ISLAM, MANJURUL M M, KIM J M. Reliable multiple combined fault diagnosis of bearings using heterogeneous feature models and multiclass support vector Machines[J]. Reliability Engineering and System Safety, 2019, 184:55-66.
[2] HAN T, ZHANG L W, YIN Z J, et al. Rolling bearing fault diagnosis with combined convolutional neural networks and support vector machine[J]. Measurement, 2021, 177, 109022.
[3] 张龙, 吴荣真. 滚动轴承性能退化评估的自适应频带熵能比指标[J]. 振动与冲击, 2021, 40(02): 63-71+110.
ZHANG L, WU R Z. Performance degradation assessment of rolling bearings based on the entropy energy ratio of adaptive frequency band[J]. Journal of Vibration and Shock, 2021, 40(02): 63-71+110.
[4] XIANG S, QIN Y, ZHU C C, et al. LSTM networks based on attention ordered neurons for gear remaining life prediction [J]. ISA transactions, 2020, 106:343-354.
[5] ZHAI X, WEI X, YANG J. A comparative study on the data-driven based prognostic approaches for RUL of Rolling Bearings[C]. In: 2019 IEEE symposium series on computational intelligence (SSCI). IEEE; 2019.
[6] QIN Y, XIANG S, CHAI Y, et al. Macroscopic-microscopic attention in LSTM networks based on fusion features for gear remaining life prediction[J]. IEEE Trans Ind Electron 2020,67(12): 10865-10875.
[7] Zhao R, Yan R, Wang J, et al. Learning to monitor machine health with convolutional Bi-directional LSTM networks [J]. Sensors. 2017, 17(2):273.
[8] CHEN D L, QIN Y, WANG Y, et al. Health indicator construction by quadratic function-based deep convolutional auto-encoder and its application into bearing RUL prediction [J]. ISA transactions, 2021, 114:44-56.
[9] QIN Y, CHEN D, XIANG S, et al. Gated dual attention unit neural networks for remaining useful life prediction of rolling bearings[J]. IEEE Trans Ind Inf, 2020, 17(9):6438-6447.
[10] LEE J, JIANG J. Time Series Multi-Channel Convolutional Neural Network for Bearing Remaining Useful Life Estimation[C]. 2019 IEEE Eurasia Conference on IOT, Communication and Engineering (ECICE), Yunlin, Taiwan, 2019, 408-410.
[11] YAN H, QIN Y, XIANG S, et al. Long-term gear life prediction based on ordered neurons LSTM neural networks[J]. Measurement, 2020, 165: 108205.
[12] LU C, WANG Z Y, ZHOU B. Intelligent fault diagnosis of rolling bearing using hierarchical convolutional network based health state classification[J]. Advanced Engineering Informatics, 2017, 32: 139-151．
[13] 张继冬,邹益胜,邓佳林,等. 基于全卷积层神经网络的轴承剩余寿命预测[J]. 中国机械工程, 2019, 30(18): 2231-2235.
ZHANG J D, ZOU Y S, DENG J L, et al. Bearing Remaining Life Prediction Based on Full Convolutional Layer Neural Networks[J], China Mechanical Engineering, 2019, 30(18): 2231-2235.
[14] 熊鹏, 汤宝平, 邓蕾, 等. 基于动态加权密集连接卷积网络的变转速行星齿轮箱故障诊断[J]. 机械工程学报, 2019, 55(07): 52-57.
XIONG P, TANG B P, DENG L, et al. Fault Diagnosis for Planetary Gearbox by Dynamically Weighted Densely Connected Convolutional Networks[J]. Journal of Mechanical Engineering, 2019, 55(07): 52-57.
[15] REN L, CUI J, SUN Y Q, et al. Multi-bearing remaining useful life collaborative prediction: a deep learning approach[J]. Journal of Manufacturing Systems, 2017, 43: 248-256．
[16] 曹正志,叶春明. 考虑转动周期的轴承剩余使用寿命预测[J]. 计算机集成制造系统, 2021, 1-14. http://kns.cnki.net/kcms/detail/11.5946.TP.20210702.1723.008.html.
CAO Z Z, YE C M. Prediction of bearing remaining useful life involving therotation period[J]. Computer Integrated Manufacturing Systems, 2021, 1-14. http://kns.cnki.net/kcms/detail/11.5946.TP.20210702.1723.008.html.
[17] GÓMEZ M J，CASTEJÓN C，GARCÍA-PRADA J C. Review of recent advances in the application of the wavelet transform to diagnose cracked rotors[J]. Algorithms, 2016, 9(1):19-32.
[18] 李宏坤, 赵长生, 周帅, 等. 基于小波包－坐标变换的滚动轴承故障特征增强方法[J]. 机械工程学报, 2011, 47(19)：74-80.
LI H K, ZHAO C S, ZHOU S, et al. Fault Feature Enhancement Method for Rolling Bearing Based on Wavelet Packet-coordinate Transformation[J]. Journal of Mechanical Engineering, 2011, 47(19): 74-80.
[19] 罗荣，田福庆，冯昌林，等. 冗余小波包改进及其在齿轮箱故障诊断中应用[J]. 机械工程学报, 2014, 50(15):82-88.
LUO R, TIAN F Q, FENG C L, et al. Improved Redundant Wavelet Packet Decomposition and Its Application in Gearbox Fault Diagnosis[J]. Journal of Mechanical Engineering, 2014, 50(15):82-88.
[20] 魏娜莎, 闫献国, 谷丰收, 等. 基于声发射的活塞环-缸套摩擦润滑特征研究[J]. 机械工程学报, 2021, 57(08): 107-114.
WEI N S, YAN X G, GU F S, et al. Research on the Characteristics of Piston Ring-liner Friction Lubrication Based on Acoustic Emission[J]. Journal of Mechanical Engineering, 2021, 57(08): 107-114.
[21] 赵志宏, 杨绍普. 基于小波包变换与样本熵的滚动轴承故障诊断[J]. 振动.测试与诊断, 2012, 32(04): 640-644+692.
ZHAO Z H, YANG S P. Roller Bearing Fault Diagnosis Based on Wavelet Packet Transform and Sample Entropy[J]. Journal of Vibration，Measurement & Diagnosis, 2012, 32(04): 640-644+692.
[22] ZHAO M H, KANG M, TANG B P, et al. Deep Residual Networks With Dynamically Weighted Wavelet Coefficients for Fault Diagnosis of Planetary Gearboxes[J]. IEEE Transactions on Industrial Electronics, 2018, 65(5):4290-4300.
[23] 王奉涛, 刘晓飞, 邓刚, 等. 基于长短期记忆网络的滚动轴承寿命预测方法[J]. 振动.测试与诊断, 2020, 40(02): 303-309+419.
WANG F T, LIU X F, DENG G, et al. Remaining Useful Life Prediction Method for Rolling Bearing Based on the Long Short-Term Memory Network[J]. Journal of Vibration，Measurement & Diagnosis, 2020, 40(02): 303-309+419.
[24] 许子非, 金江涛, 李春. 基于多尺度卷积神经网络的滚动轴承故障诊断方法[J]. 振动与冲击, 2021, 40(18): 212-220.
XU Z F, JIN J T, LI C. New method for the fault diagnosis of rolling bearings based on a multiscale convolutional neural network[J]. Journal of Vibration and Shock, 2021, 40(18): 212-220.
[25] 陈仁祥, 张勇, 杨黎霞, 等. 基于整周期数据和卷积神经网络的谐波减速器健康状态评估[J]. 仪器仪表学报, 2020, 41(02): 245-252.
CHEN R X, ZHANG Y, YANG L X, et al. Health condition assessment of harmonic reducer based on integer-period data and convolutional neural network[J]. Chinese Journal of Scientific Instrument, 2020, 41(02): 245-252.
[26] JÖRGES C, BERKENBRINK C, Stumpe Britta. Prediction and reconstruction of ocean wave heights based on bathymetric data using LSTM neural networks[J]. Ocean Engineering, 2021, 232, 109046.
[27] WANG B, LEI Y G, LI N P, et al. A hybrid prognostics approach for estimating remaining useful life of rolling element bearings[J]. IEEE Transactions on Reliability, 2020, 69(1): 401-412.
[28] SKOURT B A, HASSANI A E, MAJDA A. Mixed-Pooling-Dropout for Convolutional Neural Network Regularization[J]. Journal of King Saud University - Computer and Information Sciences, 2021, DOI: 10.1016/j.jksuci.2021.05.001
[29] AN Q L, TAO Z R, XU X W, et al. A data-driven model for milling tool remaining useful life prediction with convolutional and stacked LSTM network [J]. Measurement, 2019, 154:107461.
[30] 雷亚国, 韩天宇, 王彪, 等. XJTU-SY滚动轴承加速寿命试验数据集解读[J]. 机械工程学报, 2019, 55(16): 15-20.
LEI Y G, HAN T Y, WANG B, et al. XJTU-SY Rolling Element Bearing Accelerated Life Test Datasets: A Tutorial. Journal of Mechanical Engineering, 2019, 55(16): 15-20.
[31] LEE J, QIU H, YU G and Rexnord Technical Services. Bearing Data Set. CA: IMS, University of Cincinnati. NASA Ames Prognostics Data Repository, 2007. http://ti.arc.nasa.gov/project/prognostic-data-repository/.