Abstract

Characterizing the degradation of batteries is still a matter of ongoing research due to the diverse operating conditions at which they are submitted. For example, different current discharge rates and asymmetrical charge/discharge cycles are critical operating conditions that affect the both the performance and the lifespan. This article extends and improves a previously published methodology to estimate the degradation process of Li-ion batteries by using the Kalman Filter to estimate one of the parameters of the model. Furthermore, the Kalman Filter is then combined with a Similarity-Based-Modeling framework, which integrates information of the State-of-Charge and different discharge currents in each operating cycle to estimate the degradation process. The results are obtained using information provided by the manufacturer and also with measured data. Moreover, this model is applied to a random usage profile of an Electric-Vehicle to characterize of the degradation process of the batteries under more realistic usage conditions.