An Approach for Vehicle State Estimation Using Extended Kalman Filter

In order to meet the high cost requirement of some vehicle states measured directly in vehicle active safety control system, an approach using the Extended Kalman Filter to estimate lateral and longitudinal velocity is proposed. Firstly, a vehicle dynamic model with 3 DOF, including longitudinal, lateral and yaw motions is built with MATLAB/SIMULINK. Secondly, the vehicle state estimation algorithm by the extended Kalman state observer based on the nonlinear vehicle model is achieved and the states of longitudinal, lateral acceleration and yaw rate for the vehicle are estimated online. Finally, the estimated results are compared with the results obtained from CarSim using the same parameter to verify the practicality of the proposed method.