Health Indicators Identification of Lithium-Ion Battery From Electrochemical Impedance Spectroscopy Using Geometric Analysis

Electrochemical impedance spectrum (EIS) of lithium-ion battery changes regularly with cycling, and is an effective tool for analyzing aging. However, due to the anomalous diffusions and non-exponential effects in battery, the EIS-based model is generally identified in complex and time-consuming ways, which limits its online application. In this article, a parameter identification method for an EIS-based model is proposed using geometric analysis. By fitting the impedance spectrum at intermediate frequencies with two depressed semicircles, the parameters of the EIS-based model are directly calculated, greatly reducing the computational complexity. Compared with the traditional optimization algorithms, the indexes of the proposed method are optimal, considering running time, file size, and goodness of fit. Furthermore, six model parameters are identified based on EIS at different temperatures, states of charge (SoCs), and aging cycles, and their relationship with capacity decay is analyzed. The results show that the charge transfer (CT) resistance always has an excellent linear correlation with capacity decay, and the correlation coefficient exceeds 0.94 regardless of temperature and SoC. It can be used as a health indicator for batteries. Moreover, one fractional-order parameter has a good correlation with the capacity only at high temperatures.

Automated summary

In this article, a parameter identification method for an EIS-based model is proposed using geometric analysis. By fitting the impedance spectrum at intermediate frequencies with two depressed semicircles, the parameters of the EIS-based model are directly calculated, greatly reducing the computational complexity. The relationship between six model parameters identified based on EIS at different temperatures, states of charge (SoCs), and aging cycles, and capacity decay is analyzed. The results show that the charge transfer (CT) resistance always has an excellent linear correlation with capacity decay, and can be used as a health indicator for batteries.