Estimated temperature distribution-based state-of-health prediction via recursive least squares in electro-thermal circuit model

Recently, high-nickel technology has attracted attention to improve energy density of battery cells, and as energy density is increased, research on state management and safety of battery cells has become more important. Particularly, when a large number of battery cells are connected, different aging tendencies may occur, which directly affects performance and safety. The aging of lithium-ion battery is determined based on capacity or resistance, but it is difficult to measure directly on the real-time. Accordingly, research on soundness indicators for monitoring aging of lithium-ion batteries in real time is active. In addition, the health indicators (HIs) are also used as information for determining the performance and safety of the battery by determining aging. In this study, using two lithium-ion battery of different composition ration both Ni0.5Co0.2Mn0.3O2 (NCM523) and Ni0.6Co0.2Mn0.2O2 (NCM622) pouch-type cells were subjected to cycling tests, and the results revealed their different characteristics. Besides, in this study, useful term resistance is extracted in real time to identify aging based on the recursive least squares (RLS). And the extracted resistance is used as heat generation of the thermal model. In addition, the main parameters of the thermal model implemented for temperature estimation were verified through simulation. The thermal-electrical model can perform temperature estimation based on a real-time current-voltage input, and the resistance and temperature estimation performance due to aging is confirmed using the cycle experiment data. It was confirmed that both resistance and temperature can be used as important information for the battery state-of-health estimation.