UKF and EKF estimator design based on a nonlinear vehicle model containing UniTire model

Dynamic model based vehicle state variables observer is a step towards economical on-board sensing system. However a complex model always leads to a control system with a poor real-time performance, while a simple model cannot exhibit real characteristics of a vehicle. In order to make an accurate and sententious estimate for yaw rate and side slip angle, an ameliorated 2-DOF bicycle model containing UniTire model is introduced. Then two observers based on Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) are introduced. And validity of the two algorithms is verified by simulation test and contrast is brought out respectively. The simulation results show that the UKF based observer performs better in accuracy and computing speed.