Battery pack diagnostics for electric vehicles: Robustness of the state of health measurement and differential voltage analysis at the vehicle level

Battery aging reduces an electric vehicle’s (EV) range and power capabilities, which are key factors for purchasing (second-hand) electric vehicles. As both properties degrade with time and usage, reliable methods are necessary to track battery aging and provide insights into the degradation progression. In the battery domain, aging is often assessed through remaining energy measurements such as for the state of health (SOH) metric, and differential voltage (DV) analysis. Recent studies have demonstrated that both methods are similarly applicable at the vehicle level using low-power charging measurements. In this article, the robustness of vehicle-level charging measurements for SOH determination and DV analysis are evaluated under variations in external conditions, including charging power, ambient temperature, current direction, and resting state of charge, using the state-of-the-art EVs Volkswagen ID.3, Cupra Born and Tesla Model 3. Additionally, the reproducibility and comparability of the measurement procedure are analyzed through measurements with multiple vehicles of the same model. Overall, vehicle-level SOH determination and DV analysis prove to be robust and reproducible when boundary conditions are consistently met, e.g, containment of charging power and temperatures. The results are also comparable across identical EV models, enabling the use of pristine reference vehicles as a basis for assessing aging in EVs lacking an initial battery characterization.