A novel voltage matching-adaptive extended Kalman filtering construction method for the state of charge prediction of lithium-ion batteries

Abstract

Accurately estimating the state of charge of lithium-ion batteries is of great significance to the development of the new energy industry. This research proposes a method for estimating the state of charge of lithium-ion batteries based on a voltage matching-adaptive extended Kalman filtering algorithm. The voltage matching part and the first-order resistance-capacitance (RC) part is combined into a new equivalent circuit model. This model improves the accuracy of voltage simulation at different charging and discharging stages through segment matching. Model-based adaptive extended Kalman filter algorithm adds a noise correction factor to adaptively correct the influence of noise on the estimation process and improve the estimation accuracy. The forgetting factor is introduced to improve the real-time performance of the algorithm. To verify the reliability of the model and algorithm, a multi-condition experiment is carried out on the lithium-ion battery. The verification results show that the simulation error of the circuit model to the working state of the lithium-ion battery is less than 0.0487V. The improved algorithm can accurately estimate the state of charge of lithium-ion batteries, the estimation accuracy of the discharge stage is 98.34%, and the estimation accuracy of the charging stage is 97.75%.