Precise Vehicle Mass and Parameter Estimation for Energy Consumption Prediction with a Robust Extended Kalman Filter

With the rising share of electric cars in the automotive market, features like route optimization and range estimation become more and more important. To enable these features, the energy consumption on a planned route has to be estimated. An important part of estimating the energy consumption of electric vehicles is to determine the driving resistance parameters, namely the mass, air drag area and rolling resistance coefficient. These can vary between each trip due to loading or a different vehicle configuration, i.e. a roof box, a bicycle rack or a trailer. The main contribution of this work is the implementation and realworld validation of a new estimation algorithm, a robust extended Kalman filter (EKF), and comparison with current state-of-theart algorithms. The EKF offers superior performance compared to currently used algorithms, which are standard Kalman filters or recursive least square (RLS) algorithms. The algorithm is validated with real-world driving data from different vehicle configurations.