Online multi-scenario impedance spectra generation for batteries based on small-sample learning

• Published in: Cell reports physical science 2024-08, Vol.5 (8), p.102134, Article 102134
• Main Authors: Zhu, Jiajun, Lai, Xin, Tang, Xiaopeng, Zheng, Yuejiu, Zhang, Hengyun, Dai, Haifeng, Huang, Yunfeng
• Format: Article
• Language: English
• Subjects: data-driven; electrochemical impedance spectroscopy; fractional-order circuit model; lithium-ion batteries; virtual battery
• ISSN: 2666-3864; 2666-3864
• DOI: 10.1016/j.xcrp.2024.102134

The onboard acquisition of data from electrochemical impedance spectroscopy (EIS) is critically important to the state assessment and fault diagnosis of mobile batteries, but it is technically challenging due to the stringent test requirements, limited modeling data, and varying mechanisms among batteries with different chemistries. This paper, without requiring any additional sensors, extends the traditional EIS measurement to online generation and covers most battery-using scenarios, including different battery chemistries, aging degrees, remaining capacities, and temperatures. Virtual simulation and transfer techniques are employed to train a deep neural network with a significantly reduced dataset. Specifically, we train the network with no more than 24 groups of data and achieve an average relative error lower than 5%, outperforming most “big data”-involved algorithms of its kind. Our method lowers the threshold of using EIS onboard and unlocks new opportunities to monitor the battery’s performance in both time and frequency domain comprehensively in real time. [Display omitted] •Onboard electrochemical impedance spectroscopy estimation for batteries•Different temperatures, states of charge, and states of health are examined•Deep neural network is trained with only 24 groups of experimental data•Typical error lower than 1.5 mΩ, and technique is suitable for arbitrary dynamic load profiles This study extends traditional offline electrochemical impedance measurements to online generation of spectra, usable for arbitrary dynamic load profiles, different battery chemistries, aging degrees, measuring remaining capacities, and different temperatures. Zhu and Lai et al. lower the threshold of using impedance spectroscopy and unlock opportunities to monitor a battery’s performance in both time and frequency domains.