Vehicle lateral and longitudinal velocity estimation based on Adaptive Kalman Filter

In order to meet the requirements of lateral velocity (v_y) and longitudinal velocity (v_x) in vehicle active safety control systems, and to modify the impact of noise change on the estimation accuracy, a novel method based on Adaptive Kalman Filter (AKF) is proposed for estimation of v_y and v_x in this paper by updating the mean and covariance of noise online. This method is evaluated under a variety of driving conditions and the estimation values are compared with simulator values from CarSim. The results demonstrate that the proposed method is robust and can improve the estimation accuracy of v_y and v_x.