A low-complexity state of charge estimation method for series-connected lithium-ion battery pack used in electric vehicles

Practical and accurate state of charge estimation for battery pack is a challenging task due to the inconsistency among in-pack cells. In this paper, we propose a low-complexity state of charge estimation method for series-connected battery pack. Firstly, to reduce computation cost, the capacity and state of charge calculations of battery pack are effectively simplified based on the probability theory. Secondly, we propose a selection method for the representative cells, which is validated by the simulation under Dynamic Stress Test. Subsequently, the state of charge of each representative cell is estimated by the recursive least squares-adaptive extend Kalman filter algorithm successively. Finally, the aging experiments for LiNCM and LiFePO4 battery packs are carried out to validate the feasibility of the simplified method, and the complexity of the three estimation methods is compared. Experimental results indicate that the capacity of battery pack depends only on the representative cells at different cycles, and the proposed representative cell method can estimate the state of charge of battery pack with high accuracy and low complexity. The mean absolute errors and root mean square errors for state of charge estimation are less than 3% under Urban Dynamometer Driving Schedule test at different cycles.