Stepwise Parameter Estimation Approach for Enhanced Single Particle Model in Lithium-Ion Batteries Using Genetic Algorithm

초록

With the growing demand for electric vehicles, there is an increasing need to accurately assess the internal states of lithium-ion batteries to enhance both performance and safety. This study introduces a novel parameter estimation approach based on an enhanced single particle model, which ensures high accuracy compared to experimental data. The parameter estimation process is carried out in four stages: determination of stoichiometric values (Stage 1), model parameter estimation using genetic algorithm (GA) (Stage 2), evaluation of entropy coefficients with SOC (Stage 3), and re-estimation of model parameters (Stage 4). The optimization algorithm aims to minimize the root mean square error between the experimental data and model results, targeting voltage and temperature errors within 30 mV and 0.5 degrees C, respectively. The proposed approach is validated using three commercial cylindrical LIBs with different chemistries (NCA, NMC, and LFP as cathode materials). The comparison results under constant current discharge and US06 driving test power cycle show good accuracy for both electrochemical and thermal characteristics. A stepwise approach for ESPM parameter estimation ensures high accuracy and efficiency.Electrochemical and thermal parameters optimized using a genetic algorithm (GA).Validated using three commercial cylindrical LIBs with NCA, NCM, and LFP chemistries.Achieved high accuracy in voltage and temperature predictions under various conditions.

키워드