Transfer learning for remaining useful life prediction based on consensus self-organizing models

Fan, Yuantao; Nowaczyk, Sławomir; Rögnvaldsson, Thorsteinn

doi:10.1016/j.ress.2020.107098

Cited by 63 publications

(33 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, promoting data sharing via privacy-preserving or data monetisation can also solve data scarcity problems in some use cases of the energy sector, such as forecasting the condition of electrical grid assets (Fan et al, 2020). Moreover, combination of heterogeneous data sources (e.g., numerical, textual, categorical) is a challenging and promising avenue of future research in collaborative forecasting (Obst et al, 2019).…”

Section: Collaborative Forecasting In the Energy Sector 126mentioning

confidence: 99%

Forecasting: theory and practice

Petropoulos,

Apiletti,

Assimakopoulos

et al. 2020

Preprint

View full text Add to dashboard Cite

Forecasting has always been in the forefront of decision making and planning. The uncertainty that surrounds the future is both exciting and challenging, with individuals and organisations seeking to minimise risks and maximise utilities. The lack of a free-lunch theorem implies the need for a diverse set of forecasting methods to tackle an array of applications. This unique article provides a non-systematic review of the theory and the practice of forecasting. We offer a wide range of theoretical, state-of-the-art models, methods, principles, and approaches to prepare, produce, organise, and evaluate forecasts. We then demonstrate how such theoretical concepts are applied in a variety of real-life contexts, including operations, economics, finance, energy, environment, and social good. We do not claim that this review is an exhaustive list of methods and applications. The list was compiled based on the expertise and interests of the authors. However, we wish that our encyclopedic presentation will offer a point of reference for the rich work that has been undertaken over the last decades, with some key insights for the future of the forecasting theory and practice.

show abstract

Section: Collaborative Forecasting In the Energy Sector 126mentioning

confidence: 99%

Forecasting: theory and practice

Petropoulos,

Apiletti,

Assimakopoulos

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…The other essential line of action is to look for different DL algorithms and architectures like RNN, GAN, TL and RL. Recent works have proposed approaches based on DL to resolve the problem of anomaly detection in time-series [ 28 , 125 , 127 , 139 , 141 , 142 ]. Nevertheless, new proposals in this research line will be necessary.…”

Section: Discussionmentioning

confidence: 99%

“…Several ML algorithms and methods have been used to predict failures and RUL. Some approaches explored the use of classical algorithms as LR [ 22 ], SVR [ 23 ], SVM [ 24 ], RF [ 25 ] while osthers explored the combined use of algorithms with step phased approaches: ARIMA and SVM [ 26 ], SVR and SVM [ 27 ] and TL with RF [ 28 ]; and also with a comparative approach: RF, QRF, DT, KNN, SVR and PCR [ 25 ]. In here, we briefly review recent works used traditional ML methods in PdM applications.…”

Section: Data-driven Pdmmentioning

confidence: 99%

“…The MAE and MAPE were used to evaluate an approach for RUL estimation on two datasets was proposed and evaluated in [ 74 ], and an algorithm based on ARIMA and SVM proposed in [ 26 ] for RUL estimation. MAPE and RMSE were also used in [ 28 ] to evaluate a mapping function using RF regression model for predicting RUL of equipment under the scenario that labeled data are only available for the source domain.…”

Section: Evaluation Metrics In Pdm!mentioning

confidence: 99%

See 1 more Smart Citation

A Survey on Data-Driven Predictive Maintenance for the Railway Industry

Davari

Veloso

Costa

et al. 2021

Sensors

View full text Add to dashboard Cite

In the last few years, many works have addressed Predictive Maintenance (PdM) by the use of Machine Learning (ML) and Deep Learning (DL) solutions, especially the latter. The monitoring and logging of industrial equipment events, like temporal behavior and fault events—anomaly detection in time-series—can be obtained from records generated by sensors installed in different parts of an industrial plant. However, such progress is incipient because we still have many challenges, and the performance of applications depends on the appropriate choice of the method. This article presents a survey of existing ML and DL techniques for handling PdM in the railway industry. This survey discusses the main approaches for this specific application within a taxonomy defined by the type of task, employed methods, metrics of evaluation, the specific equipment or process, and datasets. Lastly, we conclude and outline some suggestions for future research.

show abstract

“…They introduced the idea of zero-shot learning into industrial fields and proposed a zero-sample fault diagnosis method based on the attribute transfer method. RUL prediction studies based on TL are still relatively few in number, as far as the authors know ( Fan, Nowaczyk & Rögnvaldsson, 2020 ; Mao, He & Zuo, 2020 ; Sun et al, 2019 ; Zhu, Chen & Shen, 2020 ).…”

Section: Literature Reviewmentioning

confidence: 99%

Data-driven remaining useful life prediction based on domain adaptation

Wen

Xiao

Wang

et al. 2021

PeerJ Computer Science

View full text Add to dashboard Cite

As an important part of prognostics and health management, remaining useful life (RUL) prediction can provide users and managers with system life information and improve the reliability of maintenance systems. Data-driven methods are powerful tools for RUL prediction because of their great modeling abilities. However, most current data-driven studies require large amounts of labeled training data and assume that the training data and test data follow similar distributions. In fact, the collected data are often variable due to different equipment operating conditions, fault modes, and noise distributions. As a result, the assumption that the training data and the test data obey the same distribution may not be valid. In response to the above problems, this paper proposes a data-driven framework with domain adaptability using a bidirectional gated recurrent unit (BGRU). The framework uses a domain-adversarial neural network (DANN) to implement transfer learning (TL) from the source domain to the target domain, which contains only sensor information. To verify the effectiveness of the proposed method, we analyze the IEEE PHM 2012 Challenge datasets and use them for verification. The experimental results show that the generalization ability of the model is effectively improved through the domain adaptation approach.

show abstract

Transfer learning for remaining useful life prediction based on consensus self-organizing models

Cited by 63 publications

References 36 publications

Forecasting: theory and practice

Forecasting: theory and practice

A Survey on Data-Driven Predictive Maintenance for the Railway Industry

Data-driven remaining useful life prediction based on domain adaptation

Contact Info

Product

Resources

About