Parameter Identification of Pseudo-Two-Dimensional Model for Lithium-Ion Batteries with Silicon/Graphite Composite Electrodes

Abstract

Listen

Translate article

In this study, we assess the performance of lithium-ion batteries (LIBs) equipped with silicon/graphite composite electrodes and estimate the electrochemical properties of each constituent material. We employ a composite electrode model integrated into the conventional pseudo-two-dimensional (P2D) model to simulate the electrochemical dynamics of the battery. First, the accuracy of the numerical model is validated by comparing its voltage predictions with existing experimental data. The model successfully captures the voltage characteristics of LIBs across various C-rates. Additionally, the interaction effects between silicon and graphite within the composite electrode are validated by comparing them to previous simulation results. Specifically, the reactivity of silicon and graphite, which varies with the operating voltage range, is accurately assessed under various operating conditions.Parameter estimation for the electrochemical model is performed using a genetic algorithm. Ten parameters significant to the electrochemical performance of each material are selected, considering their physical significance and the feasibility of experimental validation. These parameters are optimized by minimizing a fitness function defined by the relative errors between the experimental and simulated voltage, utilizing 1/3C discharge and pulse data from different states of charge (SOCs). The simulated voltages with the optimized parameters align well with the experimental data, accurately representing both the constant-current discharging voltage and the voltage dynamics during pulse application and removal. This alignment confirms that the estimated parameters effectively characterize the electrochemical behavior of LIBs with the composite electrode.A comparative analysis of parameter identification using only constant-current discharging data versus a combination of constant-current and pulse data reveals significant variations in the distribution of estimated values. Parameters derived solely from constant-current data inadequately fit the pulse data, highlighting the necessity of incorporating diverse data types for reliable parameter estimation.The methodology proposed in this work provides a robust framework for evaluating the electrochemical properties and performance of LIBs featuring silicon/graphite composite electrodes. This approach offers a detailed assessment of the material-level properties by identifying critical electrochemical parameters. Future work will focus on developing an efficient experimental protocol for parameter estimation, employing sensitivity analysis and case studies across different voltage profiles to ensure reliable parameter determination. Figure 1