State of Health Estimation for Lithium-Ion Batteries Based on Peak Region Feature Parameters of Incremental Capacity Curve

At present, researchers have widely used feature parameters (FPs) of incremental capacity (IC) curve to estimate the state of health (SOH) of lithium ion batteries. The FPs are commonly extracted from a whole peak in the IC curve. The method fails to consider the effect of the FPs extracted from different ranges of the peak on the accuracy of estimated SOH. In order to provide an accurate SOH estimation, we select the FPs from the peak region(∆Vreg, a state of charge range of a peak). Then SOH estimation is achieved by setting up the relationship between the SOH and the FPs based on Gaussian process (GP) regression. Results show that the accuracy of SOH estimation is sensitive to the different FPs, according to the estimated SOH under the three ∆Vreg. Furthermore, the comparison of the eleven ∆Vreg of FPs that the data come from the NASA No.5, 6, 7 and 18 batteries between 23.1% and 100% is studied. It is found that the estimated SOH root mean square error is less than 2% when the ∆Vreg of No.6,7 and 18 batteries are in the regions of [53.4%,88.1%],[50.4%,92.3%] and [42.3%,100%], respectively. It is indicated that that SOH estimation is more sensitive to the above peak region. This method gives an approach to achieve the high precision of SOH estimation because we prove that the SOH estimation is sensitive to ∆Vreg. © 2021, Electrical Technology Press Co. Ltd. All right reserved.