Estimating and Identifying Parameters from Discharge Curves of Lithium-Ion Batteries

Abstract

In order to improve the performance of Lithium ion secondary batteries (LiBs) for electric vehicles and hybrid electric vehicles, it is very important to understanding the internal transport phenomena and resistance under high rate condition to increase power density. In our previous studies, we focused on the actual porous electrode structure, and the Li ion and electron conductivity was evaluated by the effective conductive path with the actual reconstruction electrode structure [1]. In addition, the charge-discharge performance was simulated with the model of reaction and mass transfer in heterogeneous porous electrode structure [2]. In this simulation, as assumptions, Li concentration and potential distribution in active material (AM) and electrolyte (EL) were calculated with mass transport simulation and Butler-Volmer equation as the boundary condition of the interface between AM and EL. From these calculation results, the effect of heterogeneous porous structure, which consist of active material, conductive material and binder, on the cell performance could be evaluated. Then, the values of parameters in this model are essential to accuracy of it. However, some parameters cannot be measured precisely in the experiment, especially, reaction rate constant, tortuosity of separator, Li+ diffusion coefficient, and transport number in electrolyte are treated as literature value or ex-situ results. In this paper, we established the method of identifying these parameters by using experimental data in various condition and complex method [3]. We conducted fitting on experimental data of some SOC and some discharge rate condition. We adopted RMSE (Root Mean Square Error) to judge this fitting. Figure1 shows the overview of complex method. it is non-linear optimization method. Figure2 indicate transition of RMSE. From this, we can understand RMSE decreasing, and then, the value finally settle at 0.021. Thus the parameters can be identified. In addition, we will present the results of electrode structure optimization with complex method for high output power density cell.