TY - GEN
T1 - A Bayesian Network Approach for Imbalanced Fault Detection in High Speed Rail Systems
AU - Li, Yan Fu
AU - Liu, Jie
N1 - Publisher Copyright:
© 2018 IEEE.
PY - 2018/8/27
Y1 - 2018/8/27
N2 - Safety and reliability of High Speed Trains (HSTs) are crucial factors for their development as mass transport means. For this reason, they are highly monitored systems, and large amounts of data are collected and used for efficient operation and maintenance. In this paper, we focus on extracting knowledge from these data for fault detection in the braking system of HSTs. A probabilistic, explainable framework is proposed, based on an objective-oriented Bayesian Network (BN). A symmetric uncertainty-based feature selection method is combined with BN, for the first time, for reducing the dimensionality of the original data. The imbalance ratio of the data can be up to more than 300 and sensitivity analysis of the method is performed. Experiment results show that the proposed approach is more accurate than published method.
AB - Safety and reliability of High Speed Trains (HSTs) are crucial factors for their development as mass transport means. For this reason, they are highly monitored systems, and large amounts of data are collected and used for efficient operation and maintenance. In this paper, we focus on extracting knowledge from these data for fault detection in the braking system of HSTs. A probabilistic, explainable framework is proposed, based on an objective-oriented Bayesian Network (BN). A symmetric uncertainty-based feature selection method is combined with BN, for the first time, for reducing the dimensionality of the original data. The imbalance ratio of the data can be up to more than 300 and sensitivity analysis of the method is performed. Experiment results show that the proposed approach is more accurate than published method.
KW - cost-sensitive learning
KW - dimensionality reduction
KW - fault detection
KW - high speed trains
KW - knowledge extraction
KW - objective-oriented Bayesian network
UR - https://www.scopus.com/pages/publications/85062870121
U2 - 10.1109/ICPHM.2018.8448459
DO - 10.1109/ICPHM.2018.8448459
M3 - 会议稿件
AN - SCOPUS:85062870121
T3 - 2018 IEEE International Conference on Prognostics and Health Management, ICPHM 2018
BT - 2018 IEEE International Conference on Prognostics and Health Management, ICPHM 2018
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2018 IEEE International Conference on Prognostics and Health Management, ICPHM 2018
Y2 - 11 June 2018 through 13 June 2018
ER -