Design and Development of a Battery State of Health Estimation Model for Efficient Battery Monitoring Systems

Sensors

Jin Woo Choi¹  Taeg Keun Whangbo²  Hyoung Sun Choi³ 

[1] Cultural Contents Technology Institute, Gachon University, Sujeong-gu, Seongnam-si 461-701, Gyeonggi-do, Korea;Department of Computer Engineering, Gachon University, Sujeong-gu, Seongnam-si 461-701, Gyeonggi-do, Korea;Department of IT Convergence Engineering, Gachon University, Sujeong-gu, Seongnam-si 461-701, Gyeonggi-do, Korea;

关键词: uninterruptible power supply;    battery management system;    SoH;    clustering;    recurrent neural network;    LSTM;   

DOI  :  10.3390/s22124444

来源: DOAJ

【 摘 要 】

An uninterruptible power supply (UPS) is a device that can continuously supply power for a certain period when a power outage occurs. UPS devices are used by national institutions, hospitals, and servers, and are located in numerous public places that require continuous power. However, maintaining such devices in good condition requires periodic maintenance at specific time points. Efficient monitoring can currently be achieved using a battery management system (BMS). However, most BMSs are administrator-centered. If the administrator is not careful, it becomes difficult to accurately grasp the data trend of each battery cell, which in turn can lead to a leakage or heat explosion of the cell. In this study, a deep-learning-based intelligent model that can predict battery life, known as the state of health (SoH), is investigated for the efficient operation of a BMS applied to a lithium-based UPS device.

【 授权许可】

Unknown