Impact of cooling on battery cycle life based on direct current internal resistance and machine learning model prediction

Abstract

Lithium-ion batteries are a sustainable energy storage technology and are workhorses for electric vehicles (EVs) as well as stationary applications. However, the cycle life analysis of these batteries is critical to ensure their reliability and safety of operation, which has remained a significant challenge. The complex behavior of battery affects the performance parameters that are responsible for the degradation mechanisms. The focus on parameters such as discharge capacity, internal resistance, current and temperature rise correlates best in predicting the cycle life of batteries. The study proposes a new approach to investigate the cycle life by using Direct Current Internal Resistance (DCIR) and temperature rise behavior of lithium iron phosphate (LFP) and lithium nickel manganese cobalt oxides (NMC) battery cell chemistries for multiple discharge rates. The study also analyses the impact of cooling on battery cells to improve the cycle life of batteries. The experimental method is employed for battery cycling for 400 cycles to analyze the battery cell DCIR trend under test conditions, and this data is used for cycle life prediction. The experimental results reveal that the internal resistance of battery cells is affected by two parameters namely discharge rates and operating temperature conditions. The linear regression and machine learning models are used to predict the future cycle life of battery cells. The result reveals that cooling of cells during operation reduces the drift of internal resistance leading to increased cycle life of battery cells. The linear regression model predicts that applying cooling to LFP and NMC battery cells increases cycle life from 1382 cycles to 2570 cycles and from 1280 cycles to 2420 cycles, respectively. Additionally, the machine learning model predicts that applying cooling to both LFP and NMC battery cells increases cycle life from 1220 cycles to 2726 cycles and from 1270 cycles to 2708 cycles compared to the condition without cooling. The experimental results of 400 cycles show that the R² is more than 0.93 for all cases. The cycle life analysis clearly predicts that maintaining battery temperature within 20 to 50 ℃ with the help of cooling improves cycle life of batteries by 90% to 120%.