Machine Learning-Based Methodology for Fast Assessment of Battery Health Status
Global electric vehicle (EV) markets are rapidly expanding, and the efficient management of batteries has become increasingly important due to supply constraints of rare metals and other raw materials required for lithium-ion batteries. Accordingly, the reuse and recycling of used batteries from ear...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Batteries |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-0105/11/7/236 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839616501629321216 |
|---|---|
| author | Woongchul Choi |
| author_facet | Woongchul Choi |
| author_sort | Woongchul Choi |
| collection | DOAJ |
| description | Global electric vehicle (EV) markets are rapidly expanding, and the efficient management of batteries has become increasingly important due to supply constraints of rare metals and other raw materials required for lithium-ion batteries. Accordingly, the reuse and recycling of used batteries from early EVs are emerging as key solutions. This study proposes a machine learning-based approach to rapidly and reliably estimate the static capacity of used batteries. While conventional methods require significant measurement time, this study demonstrates that accurate static capacity estimation is possible using only short-term partial discharge data (6 min under 1C-rate CC conditions) by leveraging an RNN (recurrent neural network) architecture specialized for time-series data processing. The proposed model achieves high prediction accuracy, with an average RMSE of 28.439 mAh, average MSE of 808.799 mAh<sup>2</sup>, average MAE of 13.049 mAh, and average R<sup>2</sup> of 0.9993, while significantly reducing the evaluation time compared to conventional methods. This is expected to greatly enhance the efficiency and practicality of battery reuse and recycling processes. |
| format | Article |
| id | doaj-art-b5d7b4c7c9e34e9c9c91eb8aa15d6e04 |
| institution | Matheson Library |
| issn | 2313-0105 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Batteries |
| spelling | doaj-art-b5d7b4c7c9e34e9c9c91eb8aa15d6e042025-07-25T13:13:57ZengMDPI AGBatteries2313-01052025-06-0111723610.3390/batteries11070236Machine Learning-Based Methodology for Fast Assessment of Battery Health StatusWoongchul Choi0Department of Automotive Engineering, Kookmin University, Seoul 02707, Republic of KoreaGlobal electric vehicle (EV) markets are rapidly expanding, and the efficient management of batteries has become increasingly important due to supply constraints of rare metals and other raw materials required for lithium-ion batteries. Accordingly, the reuse and recycling of used batteries from early EVs are emerging as key solutions. This study proposes a machine learning-based approach to rapidly and reliably estimate the static capacity of used batteries. While conventional methods require significant measurement time, this study demonstrates that accurate static capacity estimation is possible using only short-term partial discharge data (6 min under 1C-rate CC conditions) by leveraging an RNN (recurrent neural network) architecture specialized for time-series data processing. The proposed model achieves high prediction accuracy, with an average RMSE of 28.439 mAh, average MSE of 808.799 mAh<sup>2</sup>, average MAE of 13.049 mAh, and average R<sup>2</sup> of 0.9993, while significantly reducing the evaluation time compared to conventional methods. This is expected to greatly enhance the efficiency and practicality of battery reuse and recycling processes.https://www.mdpi.com/2313-0105/11/7/236fast estimationpartial dischargevoltage responsemachine learning |
| spellingShingle | Woongchul Choi Machine Learning-Based Methodology for Fast Assessment of Battery Health Status Batteries fast estimation partial discharge voltage response machine learning |
| title | Machine Learning-Based Methodology for Fast Assessment of Battery Health Status |
| title_full | Machine Learning-Based Methodology for Fast Assessment of Battery Health Status |
| title_fullStr | Machine Learning-Based Methodology for Fast Assessment of Battery Health Status |
| title_full_unstemmed | Machine Learning-Based Methodology for Fast Assessment of Battery Health Status |
| title_short | Machine Learning-Based Methodology for Fast Assessment of Battery Health Status |
| title_sort | machine learning based methodology for fast assessment of battery health status |
| topic | fast estimation partial discharge voltage response machine learning |
| url | https://www.mdpi.com/2313-0105/11/7/236 |
| work_keys_str_mv | AT woongchulchoi machinelearningbasedmethodologyforfastassessmentofbatteryhealthstatus |