Temporal Convolutional Network with Attention Mechanism for Bearing Remaining Useful Life Prediction

Abstract

Accurate prediction of the Remaining Useful Life (RUL) of rolling bearings is crucial in industrial production, yet existing models often struggle with limited generalization capabilities due to their inability to fully process all vibration signal patterns. We introduce a novel multi-input autoregressive model to address this challenge in RUL prediction for bearings. Our approach uniquely integrates vibration signals with previously predicted Health Indicator (HI) values, employing feature fusion to output current window HI values. Through autoregressive iterations, the model attains a global receptive field, effectively overcoming the limitations in generalization. Furthermore, we innovatively incorporate a segmentation method and multiple training iterations to mitigate error accumulation in autoregressive models. Empirical evaluation on the PMH2012 dataset demonstrates that our model, compared to other backbone networks using similar autoregressive approaches, achieves significantly lower Root Mean Square Error (RMSE) and Score. Notably, it outperforms traditional autoregressive models that use label values as inputs and non-autoregressive networks, showing superior generalization abilities with a marked lead in RMSE and Score metrics.

DRLSTM-DA: a novel aeroengine remaining useful life prediction method based on deep optimization of multiple sensors information

“…R is the lifetime of a run-tofailure instance. ris set to 125, according to [33][34][35]. Figure 10 shows the process of constructing the actual RUL of a run-to failure instance.…”

Bayesian neural networks for uncertainty quantification in remaining useful life prediction of systems with sensor monitoring

“…Attention networks utilize attention mechanisms to emphasize task-relevant information in deep learning networks through enhanced adaptive feature representation. To incorporate uncertainty quantification, Wang et al 58 used a Bayesian Kernel attention network for RUL prediction in bearings. The method produced more interpretable and trustworthy results in terms of the credible intervals for the RUL, along with the mean RUL prediction, thus providing a lot more information for decision-making when compared to existing methods of implementing attention networks that ignore uncertainty information and thus lead to overconfident RUL predictions.…”

A CNN-LSTM-PSO tool wear prediction method based on multi-channel feature fusion

“…Lim Meng Lip [11] et al cropped the surface profile images of machined parts and input them into CNN networks for tool wear prediction, and the results showed that the CNN model can meet the tool wear prediction requirements with an accuracy of 98.9 % accuracy. Although these methods have been successful in predicting tool wear, it is still challenging to fully reveal the effective features present in the monitored signals due to the defects in the network structure [12] .…”