Remaining useful life prediction of bearings with two-stage LSTM

Chen, Qian; Ma, Xiaobing; Yan, Bingxin; Wang, Yanyan; Huang, Guifa

doi:10.1109/isas55863.2022.9757261

Cited by 6 publications

(4 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To prove the superiority of the prediction algorithm model in this paper, we have also carried out a comparative experiment of bearings' RUL prediction using the ON-LSTM [37], WTCN [22], a two-stage LSTM prediction model (two-stage LSTM) [35], the 1DCNN-ON-LSTM model without stage division mechanism and the 1DCNN-ON-LSTM model based on stage division mechanism in this paper. Figures 20(d .…”

Section: Prediction Experiments Of Bearings' Rulmentioning

confidence: 99%

“…Although the operation time of each bearing is often different and leads to errors in RUL prediction, we can reduce the error through the RUL prediction method of the bearing based on the division of operation stages [35]. Since there is no obvious degradation trend in the early stages of bearing operation, prediction after dividing the operation stages can help the model obtain the degradation trend of the bearing and improve the prediction accuracy [31,34].…”

Section: Remarkmentioning

confidence: 99%

“…It can be found that the predicted RUL value is similar to the corresponding real RUL value. In addition, to prove the superiority of the prediction algorithm model in this paper, we also have carried out a comparative experiment of bearings' RUL prediction using the ON-LSTM [37], weighted temporal convolution network (WTCN) [22], Transformer [38], a neural network model combining CNN with bidirectional gate recurrent units (CNN-BGRU) [39], a two-stage prediction model based on the multivariate feedback extreme learning machine (MFELM) [23], a two-stage LSTM prediction model (twostage LSTM) [35], the 1DCNN-ON-LSTM model without stage division mechanism and the 1DCNN-ON-LSTM model based on stage division mechanism in this paper. Figures 12(c) and (d) shows the curves after predicting PHM-bearing 1-3 and PHM-bearing 1-4 based on the ON-LSTM model.…”

Section: Prediction Experiments Of Phm-bearings' Rulmentioning

confidence: 99%

See 2 more Smart Citations

Operation stage division and RUL prediction of bearings based on 1DCNN-ON-LSTM

Guo,

Li,

Huang

2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Remaining useful life (RUL) prediction of bearings is significantly important to ensure reliable operation of bearings. In practice, it is routinely impossible to obtain the full life cycle degradation data of bearings that needs to be used in prediction. The accuracy of the RUL prediction of bearings is often affected by incomplete degradation data. Regarding this situation, this paper proposes a multi-sensor three-stage RUL prediction framework based on the one-dimensional convolutional ordered neuron long short-term memory (1DCNN-ON-LSTM) neural network. Firstly, 1DCNN is used to extract spatial features adaptively from multi-sensor’s data and fuse them into one-dimensional feature. Next, the unsupervised hierarchy mechanism of time series information based ON-LSTM is developed to determine the ‘initial degradation stage point’ and ‘rapid degradation stage point’ of the bearing from the one-dimensional feature. Once the signal features collected by sensors input to the model reach the degradation stage point, select the corresponding sensitive features as input and construct the 1DCNN-ON-LSTM model that performs RUL prediction after the degradation stage point to improve the prediction accuracy of the model. Based on the proposed hierarchy mechanism, the bearings’ operation process is divided into three operation stages: normal stage, initial degradation stage and rapid degradation stage. Finally, the experiments verify that the proposed method can effectively divide the operation stages of bearings to predict the RUL and improve the generalization ability and prediction accuracy of the model.

show abstract

Section: Prediction Experiments Of Bearings' Rulmentioning

confidence: 99%

Section: Remarkmentioning

confidence: 99%

Section: Prediction Experiments Of Phm-bearings' Rulmentioning

confidence: 99%

See 1 more Smart Citation

Operation stage division and RUL prediction of bearings based on 1DCNN-ON-LSTM

Guo,

Li,

Huang

2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…Gao et al [18] proposed a model which combined graph convolutional network (GCN) and Bi-LSTM, and after the correlation of the characteristic parameters of the bearing vibration signal was fully extracted by GCN, then using the Bi-LSTM network structure to determine the current state of the bearing. Chen et al [19] divided the LSTM into two stages before and after degradation points, obtaining different degradation features and using them as LSTM inputs to achieve higher prediction accuracy. Wang et al [20] proposed a method that decomposes fused features on multiple scales and removes high-frequency components to reduce volatility, then combines stacked self-encoders with Bi-LSTM models to achieve high-quality feature extraction and satisfied precision.…”

Section: Introductionmentioning

confidence: 99%

RUL prediction method for rolling bearing using convolutional denoising autoencoder and bidirectional LSTM

Yao,

Zhu,

Jiang

et al. 2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Remaining useful life (RUL) prediction of rolling bearing plays an important role in maintaining the safety of the equipment. However, the data collected from industrial scene often contains noises, which affects the RUL prediction precision of rolling bearing. To overcome the above problem, a data-driven scheme for RUL prediction of rolling bearing is proposed based on convolutional denoising autoencoder (CDAE) and bidirectional long short-term memory network (Bi-LSTM). In the proposed method, the vibration signal is directly used as input of the prognostics network model. Then, a denoising network model based on CDAE is built to reduce the effect of noise. Through stacking the convolutional autoencoder, the noise component is automatically removed from the raw data. Finally, the network model based on Bi-LSTM is established to extract the high-dimensional degradation characteristics of bearing and estimate the RUL of the rolling bearing. The experimental results on the Xi’an Jiaotong University bearing dataset show that the proposed method has satisfied performance of RUL prediction.

show abstract