On-board state of health estimation of LiFePO4 battery pack through differential voltage analysis

Battery state of health (SOH) estimation is a crucial challenge for electric vehicles due to the complicated aging mechanism of a battery. In this paper, a new method for battery SOH estimation is proposed using the location interval between two inflection points or the transformation parameter of the differential voltage (DV) curve. And a new algorithm to obtain the DV curve according to the center least squares method is presented. Four LiFePO4 battery cells in different aging states are tested to validate the proposed method. Results show that the linear regression of the location interval between two inflection points versus the battery capacity from one single battery cell is able to evaluate the SOH of other three battery cells within 2% error bound. Transforming the measured DV curve to be congruent with the initial DV curve provided by the manufacturer, battery cell SOH can be estimated according to the transformation parameter with the error being within 2.5%. Furthermore, a battery pack SOH monitoring framework is inferred from the inflection point location and the transformation parameter of the DV curve to realize battery pack SOH estimation.