Abstract
Recently, deep learning is widely used in the field of remaining useful life (RUL) prediction. Among various deep learning technologies, recurrent neural network (RNN) and its variant, e.g., long short-term memory (LSTM) network, are gaining more attention because of their capability of capturing temporal dependence. Although the existing RNN-based approaches have demonstrated their RUL prediction effectiveness, they still suffer from the following two limitations: 1) it is difficult for RNN to extract directly degradation features from original monitoring data, and 2) most of the RNN-based prognostics methods are unable to quantify the uncertainty of prediction results. To address the above limitations, this paper proposes a new method named Residual convolution LSTM (RC-LSTM) network. In RC-LSTM, a new ResNet-based convolution LSTM (Res-ConvLSTM) layer is stacked with convolution LSTM (ConvLSTM) layer to extract degradation representations from monitoring data. Then, predicated on the RUL following a normal distribution, an appropriate output layer is constructed to quantify the uncertainty of the forecast result. Finally, the effectiveness and superiority of RC-LSTM is verified using monitoring data from accelerated degradation tests of rolling element bearings.
Highlights
Remaining useful life (RUL) prediction, as the effective tool in reducing unplanned shutdowns caused by mechanical failures, is widely utilized in modernized industries to ensure the safety of machines and improve the production efficiency[1][2]
Among various deep learning technologies, recurrent neural network (RNN)[11][12] and its variant, e.g., long short-term memory (LSTM) network[1315], are able to effectively capture the time dependence hidden in the degradation process, and have become the promising tool in RUL prediction
To address the above-mentioned limitations, this paper proposes a new prognostic method named residual convolution long short-term memory (RC-LSTM) network for the RUL predictions of machines
Summary
Remaining useful life (RUL) prediction, as the effective tool in reducing unplanned shutdowns caused by mechanical failures, is widely utilized in modernized industries to ensure the safety of machines and improve the production efficiency[1][2]. Compared with model-based methods, data-driven methods can adaptively model the degradation process without clear physical failure mechanisms through the utilization of machine learning (ML) technologies, such as support vector machine[4] (SVM), gaussian process regression[5] (GPR), artificial neural networks[6] (ANNs), etc. Among various deep learning technologies, recurrent neural network (RNN)[11][12] and its variant, e.g., long short-term memory (LSTM) network[1315], are able to effectively capture the time dependence hidden in the degradation process, and have become the promising tool in RUL prediction. To address the above-mentioned limitations, this paper proposes a new prognostic method named residual convolution long short-term memory (RC-LSTM) network for the RUL predictions of machines. ConvLSTM) layer, which is improved from convolution long short-term memory (ConvLSTM) network[16], is first built to extract directly degradation representations and capture time dependence information from monitoring data. The superiority of the proposed ResConvLSTM is validated using vibration data from accelerated degradation tests of rolling element bearings
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have