Article Details

Abstract

Aiming at the fault of DC power loss in traction and auxiliary systems caused by abnormal temperature of nickel-cadmium alkaline battery during neutral-section passing operation of EMU, this paper takes the abnormal temperature faults of batteries at 1st position of car 3, car 5, and car 7 in a fault case of CRH400AF EMU as the research object. Various factors strongly related to the rise of battery temperature are analyzed, and multiple abnormal temperature modes are defined. Based on this, a battery voltage trend diagnosis model based on Long Short-Term Memory (LSTM) network is established, which includes three stages: data preprocessing, model construction and training, and parameter optimization and evaluation. After the model is established, the battery temperature is tracked and analyzed through the Prognostics and Health Management (PHM) system of EMU, and the high temperature fault of alkaline batteries in car 3 is accurately determined. Meanwhile, the data from Data Recorder with Wireless Transmission Device (DRWTD) is downloaded for verification and analysis, which proves the effectiveness of the graded response strategy. This method effectively solves the problem of abnormal high temperature fault of battery and has certain engineering application value.

Keywords

CRH400AF high-speed train set; LSTM; health management system; abnormal temperature; cadmium-nickel alkaline battery

References

1. [1] Li Jing, Li Jigang, Sun Dongdong, etc The current application status and development direction of high-speed train batteries [J]. Railway Vehicles, 2023, 61 (1): 1-6
2. [2] Ding Longhai, Shi Xiaoli, Liu Ming, etc Exploration of the Influence of High speed EMU Operating Conditions on Battery Performance [J]. Battery Industry, 2018, 22 (6): 283-286
3. [3] Bloomberg Research on Charging Control of Chinese Standard EMU Battery [D]. Beijing: Beijing Jiaotong University, 2020
4. [4] CHEN M Z，MA G J，LIU W B，et al. An Overview of Data-driven Battery Health Estimation Technology for Battery Management System [J].Neurocomputing，2023，532：152-169.
5. [5] YU S H，DUAN X T， WANG X Y，et al. Revolutionizing Batteries Based on Digital Twin Through AI-simulation Synergy for Design, Manufacturing, Operation, and Recycling [J/OL].National Science Open，2025，4（6）：20250054[2025-12-27].https://doi.org/10.1360/nso/20250054.
6. [6] Fan Xiaojie, Chen Zhen'an Satellite anomaly detection method based on LSTM and multiple anomaly definitions [J]. Measurement and Control Technology, 2021, 40 (11): 78-87, 95
7. [7] Li Hui, He Jing, Cheng Fuqiang, etc Anomaly detection method for satellite power system based on LSTM model [J]. Journal of Armored Forces Engineering College, 2019, 33 (3): 90-96
8. [8] HUANG H，BOYER A，DHIA S B. Analysis and Modeling of Passive Device Degradation for a Long-term Electromagnetic Emission Study of a DC-DC Converter [J].Microelectronics Reliability，2015，55（9-10）：2061-2066.
9. [9] Li Yan, Zhang Yu, Zhou Junwei Research on Fault Prediction and Health Management System Scheme for High speed Train [J]. Railway Computer Application, 2018, 27 (9): 1-7
10. [10] Wang Zhongyao, Ma Jingwen, Wang Lianfu, etc Research on the architecture of fault prediction and health management system for high-speed trains [J]. Smart Rail Transit, 2022, 59 (2): 20-23

Funding

2025 Gansu Province College Student Innovation Training Program Project "Research on Grading Response Strategy for Warning and Alarm of Cadmium Nickel Alkaline Battery Overtemperature in High Speed Train" (S202516209013); 2024 University Teacher Innovation Fund Project "CRH High speed EMU Bogie Axle Box Node Fully Automated Assembly Device" (2024A-236); 2024 Teaching Research Project "Research on Enhancing Innovation Ability of Vocational Undergraduate Students majoring in Electrical Engineering and Automation" (2024JY-03); 2024 Scientific Research Project "Feasibility Prediction of Overtaking between Connected Vehicles and Traditional Vehicles in Hybrid Scenarios" (2024KY-22)

About the authors

Li Kaijun (July 2006-), male, Han ethnicity, from Wuwei, Gansu Province, undergraduate student, research direction: electrical engineering and automation; Corresponding author: Guo Zhilong (1989.07-), male, Han ethnicity, from Tianshui, Gansu, engineer, master's student, research direction: multi-agent collaborative control, equipment fault detection and diagnosis.