A comparative study of modeling and parameter identification for lithium-ion batteries in energy storage systems

Abstract

Accurate battery model and parameter identification are crucial for battery management. Many modeling and parameter identification methods have recently been developed for lithium-ion batteries (LIBs). However, more research is required to compare the performance of these methods quantitatively under the same conditions. This work summarizes and compares parameter identification and battery modeling methods, focusing on the integer and fractional-order models. Online and offline parameter identification methods are discussed and compared with different modeling methods, including the equivalent circuit models and fractional order models. Firstly, four integer-order models (IOM) and three fractional-order models (FOM) are constructed, and the differences in model discretization are explored. Then, the effects of open circuit voltage (OCV)- state of charge (SOC) testing methods are discussed by comparing the low-current and incremental OCV-SOC tests. Secondly, a comparison is made between the representative online and offline parameter identification methods, including methods based on recursive least squares, extended Kalman filter, and metaheuristic algorithms. Finally, the performance of modeling and parameter identification methods is analyzed, focusing on their accuracy, complexity, and computational burden.