On Experiment Design for Parameter Estimation of Equivalent-Circuit Battery Models

Using Li-ion batteries in applications, such as E-bikes, requires proper battery management. In battery management, model-based state estimation techniques can be used to estimate the State-of-Charge, for which it is common to consider an Equivalent Circuit Model (ECM). Accurate model parameters are necessary to ensure a certain quality of the state estimate. The ECM parameters highly depend on the experiment used to determine them and different choices of these experiments can be found in the literature. In this paper, we investigate the experiment design for parameter estimation both quantitatively and qualitatively. The use of pulsed currents for parameter estimation, which is a commonly used experiment, is compared to using data from a road test with the E-bike. The results quantify how much the state estimation improves when the parameters are estimated using data that represent the intended application.