Trip-synchronous parameter estimation of vehicle and tire model parameters as virtual sensor for load-sensitive lightweight vehicles

Vehicle dynamics controls, like yaw rate controls, need accurate values for vehicle inertial and tire parameters. Normally those can be assumed to remain nearly constant for everyday car trips, but looking at vehicles with very low curb weights, these parameters can change on a wide range due to different passenger or luggage loads. This effect is analyzed with several load scenarios. A Kalman filter based algorithm is presented that is able to determine all vehicle and tire parameters with standard sensors during random everyday trips and within short time. Therefore, an extended nonlinear vehicle model is defined that is able to represent vehicle behavior for everyday driving profiles very well. The estimator is validated using real-world steering and velocity profiles.