Impedance-based capacity estimation for lithium-ion batteries using generative adversarial network

This paper proposes a fully unsupervised methodology for the reliable extraction of latent variables representing characteristics lithium-ion batteries (LIBs) from electrochemical impedance spectroscopy (EIS) data using information maximizing generative adversarial networks. Meaningful representations can be obtained EIS even when measured with direct current and without relaxation, which are difficult to express circuit models. The extracted were investigated as capacity degradation progressed used estimate discharge by employing Gaussian process regression. proposed method was validated under various conditions during charging discharging. results indicate that model provides more robust estimations than EIS, where mean absolute error root square less 1.74 mAh 1.87 mAh, respectively, all operating coin cells nominal 45 mAh. We demonstrate relaxation reliably represent LIBs.