Minjun Park

State-of-health estimation algorithm of Li-ion battery using impedance at low sampling rate

In this paper, an algorithm which estimates state-of-health of Li-ion battery based on impedance at low sampling rate is proposed. The algorithm includes a battery model with Warburg impedance instead of RC parallel circuit and recursive least square method for estimating the parameters. The parameters of the Warburg model and 1st order RC model at various aging cycles are analyzed. Compared with pure resistance that has similar values for the both models, the estimated Warburg impedance has more consistent values than the RC circuit parameters. The proposed algorithm can contribute to determine aging status using the estimated Warburg impedance under resistance bias by an external resistance.

Grey prediction method for forecasting the capacity of lithium-ion batteries

The battery performance is gradually decreased over its cycle life as a result of capacity loss. This paper proposes a capacity forecasting method of lithium-ion battery pack using a grey predictor. In this method, the forecasted capacity of battery pack is obtained from the previous four measured capacity data. The prediction performance of the proposed method is verified through a comparison with the prediction results from the curve-fitting based method. The experimental results show that the proposed method can accurately predict a one-step-ahead capacity value even with a small number of data.

Robust observer for state-of-charge estimation of li-ion battery with uncertainties

We developed a robust state-of-charge (SOC) estimation algorithm which considers uncertainties of matrix including internal resistance for Li-Ion battery. We used a linear matrix inequality (LMI) to acquire gain of the observer for estimating SOC. The algorithm is less accurate than estimation results of other algorithms, but has simple and fast calculation by using a time-invariant observer gain. This algorithm can contribute to acquire SOC of old battery cells which have the higher internal resistance and uncertainty of the initial battery model.