Entropy-induced temperature variation as a new indicator for state of health estimation of lithium-ion cells

Abstract

Conventional state of health (SoH) estimation is often based on the capacity or resistance of lithium-ion cells. However, online capacity or resistance measurements are rarely achievable in electric vehicles. An alternative indicator should replace capacity and resistance to make online SoH diagnosis more manageable.In this paper, we introduce a novel SoH indicator, which is extracted from the cell temperature variation rate curve during the constant current charge process. There are two observable cooling areas in the temperature variation rate curve. The distance between these two areas, defined as tmin, could serve as a gauge for cell SoH if electrode active material loss is negligible during the operation. In order to validate the assumption, test data from both calendar aging and cycle life tests on a commercial lithium-ion pouch cell is utilized. From the 544 aging data units, a linear correlation is observed between tmin and SoH with a product-moment correlation coefficient of 0.954. The same correlation is found for all the tested cells under different aging conditions, which makes tmin a promising indicator for SoH estimation.