Vehicle State Estimation Based on Unscented Kalman State Estimation

This paper presents a method of estimating vehicle states using an Unscented Kalman filter (UKF). The UKF developed estimates Vehicle motion, such as yaw rate and side slip angle, from the noisy measurement set. The vehicle state estimation using a non-linear vehicle model with Unitire tire model will be compared to the measured state which is subjected to the same tests, in order to validate the estimated state. In this paper we also discuss the estimation algorithm of UKF. The accuracy of the estimator will be tested. The ultimate aim of this work is to provide a new way of vehicle state estimation to a controller such as ESP or VDC. The result is shown that this application of the UKF is effective in estimation of vehicle state under ISO slalom and ISO double lane change conditions.