An Improved Differential Evolution for Offline Parameter Estimation in Lithium-Ion Batteries with Real-World Usage Data

Abstract

Offline parameter extraction for state-of-charge estimation (SoC) in a BMS setup for Li-ion cells (cell level and pack level) is realized using an improved differential evolution optimization algorithm. Battery modeling is performed using a 1-RC Thevenin's equivalent circuit model (ECM). Extended Kalman Filter (EKF) is used in conjunction with the proposed IDE algorithm to validate its efficiency at SoC estimation for cell-level Hybrid Pulse Power Characterization (HPPC) test and real-world driving test data. Compared to the standard HPPC method, recursive least squares estimation (RLSE), and the classical DE methods, the proposed method is capable of estimating the offline parameters accurately with lower RMSE for the EKF-based SoC estimation. © 2023 IEEE.