Deep learning to estimate lithium-ion battery state of health without additional degradation experiments

• Published in: Nature communications 2023-05, Vol.14 (1), p.2760-2760, Article 2760
• Main Authors: Lu, Jiahuan, Xiong, Rui, Tian, Jinpeng, Wang, Chenxu, Sun, Fengchun
• Format: Article
• Language: English
• Subjects: 639/166/987; 639/4077/2790; 639/4077/4079/891; Algorithms; Artificial neural networks; Deep learning; Degradation; Errors; Humanities and Social Sciences; Labels; Lithium; Lithium-ion batteries; Machine learning; multidisciplinary; Neural networks; Power management; Rechargeable batteries; Science; Science (multidisciplinary)
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-023-38458-w

State of health is a critical state which evaluates the degradation level of batteries. However, it cannot be measured directly but requires estimation. While accurate state of health estimation has progressed markedly, the time- and resource-consuming degradation experiments to generate target battery labels hinder the development of state of health estimation methods. In this article, we design a deep-learning framework to enable the estimation of battery state of health in the absence of target battery labels. This framework integrates a swarm of deep neural networks equipped with domain adaptation to produce accurate estimation. We employ 65 commercial batteries from 5 different manufacturers to generate 71,588 samples for cross-validation. The validation results indicate that the proposed framework can ensure absolute errors of less than 3% for 89.4% of samples (less than 5% for 98.9% of samples), with a maximum absolute error of less than 8.87% in the absence of target labels. This work emphasizes the power of deep learning in precluding degradation experiments and highlights the promise of rapid development of battery management algorithms for new-generation batteries using only previous experimental data. Estimation of Li-ion battery state of health is crucial but requires time- and resource-consuming degradation tests for development. Here, authors propose a deep-learning method that enables accurate estimations without additional tests, ensuring absolute errors of less than 3% for 89.4% of samples.