A mechanism identification model based state-of-health diagnosis of lithium-ion batteries for energy storage applications

Advanced lithium-ion battery systems, in multi-cell configurations and larger-scale operations, are being currently developed for energy storage applications. Furthermore, the retired batteries are being increasingly second utilized in energy storage scenes. Thus, realistic and accurate battery state of health diagnosis and related aging mechanisms identification is desired to improve the battery management and control, and eventually guarantee the reliability and safety of the battery system. A half-cell model based battery state of health diagnostic method is proposed to investigate the aging mechanisms and possible attribute to the capacity fade in a quantitative manner. Using particle swarm optimization algorithm, the half-cell model is parameterized to quantify the battery degradation mechanisms derived from the parameter variations, which describe the electrode behavior with proper matching ratio and their evolutions at different battery aging levels. The reliability and robustness of the approach has been verified and evaluated by the database of the cells experienced different aging paths. Our approach is a data-model fusion method to offer the benefits of wide applicability to various cell chemistries and operating modes.