Life-Cycle State of Charge Estimation for Lithium-ion Battery Considering Coulomb Efficiency and Capacity Decay

Accurate state of charge (SoC) estimation of lithium-ion batteries has always been a challenge over a wide life scale. In this paper, we proposed a SoC estimation method considering Coulomb efficiency (CE) and capacity decay. Health factors are extracted from a simplified electrochemical model, and show good correlation with capacity and CE. The life-cycle data from six batteries is used to train the capacity and CE models, and these of the other two ones are estimated at different aging cycles. The mean absolute errors of the capacity and CE estimates are not more than 0.1 Ah and 0.25%, respectively. Further, the life-cycle SoC estimation results are compared in four cases regarding whether to update the capacity or CE. The results show that the capacity update can significantly improve the accuracy of SoC estimation, and the additional CE update makes the accuracy even higher on a small scale. The mean absolute errors of SoC estimation are no more than 1% at different aging cycles when both are updated.