An Improved Model for Parameters Identification of Lithium-ion Battery Based on Dual Kalman Filter

Reliable model parameters identification is the key evaluation index for battery management system (BMS) in electric vehicles (EVs). To ensure the sustainability of lithium-ion battery (LIB) under unknown measurement noise, an effective LIB model with updated parameters should be developed. To soften the impact of measurement noise from the transducer, a combined equivalent circuit model (ECM) that considers the current noise as a compensation factor is introduced into the LIB. To identify the model parameters recursively based on suppression of the parameters perturbations in the ECM, a dual extended kalman filter algorithm is applied. Finally, the Dynamic Stress Test sequence (DST) and the Federal Urban Driving Schedule (FUDS) are loaded on LIB to test the validity of the improved approach. The experiment results demonstrate the effectiveness of improved model and filtering method in terms of parameters identification.