Computational Technology Resources - CCC

Y.X. Li^1,2, S.C. Xie^1,2 and R.D. Liu^1,2

¹Key Laboratory of Traffic Safety on Track, Ministry of Education, Central South University, Changsha, China
²Joint International Research Laboratory of Key Technology for Rail Traffic Safety, Central South University, Changsha, China

doi:10.4203/ccc.1.27.21

Y.X. Li, S.C. Xie, R.D. Liu, "Hybrid Feature Selection and Fault Identification of Train Bearings Based on Integrated Learning", in J. Pombo, (Editor), "Proceedings of the Fifth International Conference on Railway Technology: Research, Development and Maintenance", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 1, Paper 27.21, 2022, doi:10.4203/ccc.1.27.21

Keywords: train bearing fault diagnosis, integrated learning, feature selection, XGBoost.

Abstract

This fast and accurate implementation of train bearing fault identification has been one of the key tasks of intelligent train health maintenance. In recent years, with the development of deep network technology, some bearing fault identification solutions based on deep learning have shown strong competitiveness. However, in the process of actual train application and maintenance, it has higher requirements for data volume and more complicated calculation. Therefore, an integrated learning-based bearing fault identification scheme is proposed. Overlapping sampling is performed considering the correlation before and after time series. Feature sets are constructed based on the characteristics of train fault signals and improved based on the XGBoost (Extreme Gradient Boosting) algorithm to achieve adaptive hybrid feature selection as well as fault identification. The effectiveness and superiority over the above method is verified by testing and evaluating two open-sources bearing datasets and laboratory bearing datasets.

download the full-text of this paper (PDF, 9 pages, 890 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCC CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Conferences ISSN 2753-3239 CCC: 1 PROCEEDINGS OF THE FIFTH INTERNATIONAL CONFERENCE ON RAILWAY TECHNOLOGY: RESEARCH, DEVELOPMENT AND MAINTENANCE Edited by: J. Pombo Paper 27.21 Hybrid Feature Selection and Fault Identification of Train Bearings Based on Integrated Learning Y.X. Li^1,2, S.C. Xie^1,2 and R.D. Liu^1,2 ¹Key Laboratory of Traffic Safety on Track, Ministry of Education, Central South University, Changsha, China ²Joint International Research Laboratory of Key Technology for Rail Traffic Safety, Central South University, Changsha, China doi:10.4203/ccc.1.27.21 Full Bibliographic Reference for this paper Y.X. Li, S.C. Xie, R.D. Liu, "Hybrid Feature Selection and Fault Identification of Train Bearings Based on Integrated Learning", in J. Pombo, (Editor), "Proceedings of the Fifth International Conference on Railway Technology: Research, Development and Maintenance", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 1, Paper 27.21, 2022, doi:10.4203/ccc.1.27.21 Keywords: train bearing fault diagnosis, integrated learning, feature selection, XGBoost. Abstract This fast and accurate implementation of train bearing fault identification has been one of the key tasks of intelligent train health maintenance. In recent years, with the development of deep network technology, some bearing fault identification solutions based on deep learning have shown strong competitiveness. However, in the process of actual train application and maintenance, it has higher requirements for data volume and more complicated calculation. Therefore, an integrated learning-based bearing fault identification scheme is proposed. Overlapping sampling is performed considering the correlation before and after time series. Feature sets are constructed based on the characteristics of train fault signals and improved based on the XGBoost (Extreme Gradient Boosting) algorithm to achieve adaptive hybrid feature selection as well as fault identification. The effectiveness and superiority over the above method is verified by testing and evaluating two open-sources bearing datasets and laboratory bearing datasets. download the full-text of this paper (PDF, 9 pages, 890 Kb) go to the previous paper go to the next paper return to the table of contents return to the volume description
Back to top	©Civil-Comp Limited 2023 - terms & conditions