Precise rocket strike, requires trajectory correction according to trajectory parameters.In this paper, double aerodynamic parameter identification algorithm was used to remove appears in initial phase of signal outliers, reduce impact on subsequent Kalman filter.Intermediate outlier measurement information was used to judge based on prediction error of the Kalman filter combined with particle trajectory model for construction of the Kalman filter trajectory model.The Kalman filter trajectory model was used to remove and filter outliers of a section of flight parameters, and extrapolate the rockets impact point.The simulation results show that, the filter can remove and filter the outliers in radar measurement data effectively.We can know that prediction accuracy can increase by postponing extrapolation.