A novel method for journal bearing degradation evaluation and remaining useful life prediction under different working conditions. (June 2021)

• Record Type: Journal Article
• Authors: Ding, Ning; Li, Hulin; Yin, Zhongwei; Jiang, Fangmin
• Is Part Of: Measurement. Volume 177(2021)
• Publication Date: 2021-06
• Subjects: Journal bearing -- Degradation performance evaluation -- RUL prediction -- Domain adaptation -- Transfer learning; RUL Remaining useful life -- LSTM Long Short-Term Memory -- DNN Deep Neural Network -- FCM Fuzzy c-means -- DOT Degradation occurrence time -- AE Auto-encoder -- SAE Stacked Auto-encoder -- DBN Deep Belief Network -- CNN Convolutional Neural Network -- RNN Recurrent Neural Network -- TCA Transfer component analysis -- MMD Maximum mean discrepancy -- COVAL Covariance alignment -- JDA Joint distribution adaptation -- DANN Domain Adversarial Neural Network -- MLP Multiple Layer Perceptron -- RKHS Reproducing Kernel Hilbert Space -- RBF Radial basis function -- MSE Mean square error -- SGD Stochastic Gradient Descent -- MAE Mean absolute error -- RMSE Root mean square error -- FC Frequency center -- RMSF Root mean square frequency -- RVF Root variance frequency -- PC1 First principal component -- PCA Principal component analysis; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2021.109273
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: Journal bearing degradation stage division is implemented by the FCM algorithm. A novel transfer learning method for RUL prediction is proposed. Journal bearing RUL prediction with different working conditions is implemented. Abstract: Accurate bearing degradation performance analysis and remaining useful life (RUL) prediction are significant to prevent major accidents and economic losses in industry. Data-driven methods have emerged as reliable algorithms for RUL prediction. These existing approaches assume that the training (source) and testing (target) samples have the same probability distribution. However, the obtained run-to-failure datasets with different work conditions are usually from different domains. The distribution discrepancy between the source and target domain will reduce the accuracy of RUL prediction models when only source domain data in one working condition is trained. To solve the problem, this paper proposes a novel transfer learning method for journal bearing RUL prediction under different work conditions based on the LSTM-DNN network with domain adaptation. The multi-sensor run-to-failure datasets of journal bearings are collected and the extracted multi-sensor features are used for degradation assessment through the fuzzy c-means (FCM) clustering algorithm and the determination of degradation occurrence time (DOT). The multi-sensor feature representations and RUL values after DOT are used to validate the effectiveness of the proposed model. The results show that the proposed method has higher accuracy in journal bearing RUL prediction under different work conditions and outperforms other transfer learning approaches.