Automatic Estimation of Parameters of a Reduced Order Electrochemical Model for Lithium-Ion Batteries at the Beginning-of-Life

摘要

This paper presents a procedure for automatic estimation of parameters of a lithium-ion battery at the beginning-of-life. Based on a reduced order electrochemical model, a two-step sensitivity analysis is employed to group the parameters and divide them using the best state-of-charge (SOC) window. Then, a genetic algorithm is applied to minimize a multi-objective function that is the L2 norm of the voltage error for each SOC window. Results of the procedure are validated against a multiple cycle and an actual driving cycle test, which keep the absolute voltage errors less than 12mV.