Vehicle Trajectory Prediction Method Based on “Current” Statistical Model and Cubature Kalman Filter

Vehicle motion trajectory prediction is the basis of vehicle collision early warning or conflict resolution. In order to improve accuracy prediction, a method based on “current” statistical (CS) model and cubature Kalman filter (CKF) proposed. This considers acceleration variation rules in actual process state equation, so that estimated value can be consistent with real range. condition overcomes limitation general model, which ignores change, it improved accuracy. addition, this also avoids large amount computational resources required, being some new methods describe fluctuations. The at intersection crossed by Yingbin Avenue Qiche Nanchang selected verify tracking performance Constant Acceleration-Unscented Filter (CA-UKF), Current Statistical-Unscented (CS-UKF), CS-CKF models. results show has superior effectiveness than CA-UKF CS-UKF models, improves prediction.