Parameter identification and sensitivity analysis of lithium-ion battery via whale optimization algorithm

Abstract The reaction mechanism of lithium-ion batteries is directly affect system safety and performance. Understanding the actual battery status through model simulation has become an important issue in battery management systems. This research proposes a non-destructive parameter identification method that uses whale optimization algorithm with unique global searching to identify the parameters of the electrochemical model and analyze the sensitivity of important parameters in the battery model. First, we establish an experimental platform and conduct four conditions, including 1C, 0.5C, 2C and one driving cycle. Moreover, 1C charge and discharge are taken as the benchmark for the parameter identification. After obtaining the key parameters of the battery, the battery performance prediction is carried out for the remaining three types. The terminal voltage of 0.5C, 2C and the road driving prediction errors are less than 15.45 mV, and the battery SOC errors are less than 1.21%. The results of parameter sensitivity studies show that the electrode-related parameters for the migration of lithium ions are the key to calculating the accuracy of the performance, and the porosity of the electrode has the greatest influence. The accurate parameter identification and the sensitivity provide a future methodology to design a proper battery model for battery management systems.