Improved SOC Estimation for Lithium-Ion Cells Valid for Different Temperatures and States-of-Health

Lithium-ion batteries are the most favored energy storage technology for high-efficiency hybrid and electrical vehicles. Online State-of-Charge (SOC) estimation is required for this application to estimate the remaining cruising distance. However, variation of battery parameters with temperature and cycle life has to be taken into account in order to achieve high accuracy.

In this work electrical tests on lithium-ion cells in different States-of-Health are performed and used to extract model parameters such as open-circuit-voltage and impedance. High precision test equipment has been developed to accurately track the true SOC of the cell during measurements. A strong influence of cycle-life on electrical battery behavior is observed.

A dynamic cell model based on the measurement results including temperature and aging effects is generated and subsequently used for SOC estimation with an Extended-Kalman-Filter. The results show high accordance of the estimated SOC with reference values, and significant improvements of estimation accuracy are shown for cells with high aging.