Battery State Estimation for High Power Safety Critical Settings with Application to eVTOL Aircraft

In electric vertical takeoff and landing (eVTOL) aircraft applications, a major function of the battery management system (BMS) is to provide estimates of battery state of power (SOP) to the pilot. These algorithms are predictive in nature and their accuracy relies on the accuracy of the underlying predictive model and the initial conditions. Here we focus on how the initialization, often provided through a state of charge (SOC) estimator can profoundly impact SOP estimation. We use the extended Kalman filter (EKF) framework to design an estimator for a recently proposed equivalent circuit model for eVTOL applications. We demonstrate that in contrast with electric vehicles, high power applications such as eVTOL are clearly influenced by internal states beyond just the SOC and that a holistic approach is needed in designing the state estimator to ensure satisfactory performance. Evaluation on over 1000 experimental eVTOL flight profiles underlines the utility of the proposed state estimation in reducing the uncertainty in an application-specific SOP prediction error by more than 60\% compared to open-loop predictions.