Due to the mismatch between the tracking model and the real motion state of the target， the estimation accuracy of the high order motion feature of the target is poor in single-station radar.In this paper， the estimation accuracy of high order motion feature of the target is improved from the perspective of radar networking and radial velocity augmentation measurement. The sequential unscented Kalman filter algorithm is used to centrally fuse the asynchronous measurement information of multiple radars. The estimation effect of the high order motion feature of the target under different ranging， angular accuracy and radial velocity augmentation measurement is analyzed， and the influence of the radar station layout on the estimation effect of the high order motion feature of the target is analyzed by using the geometric accuracy factor of double sensors to optimize the radar station layout. Through simulation experiment， it is found that the radar ranging accuracy in this radar networking mode has a greater impact on the estimation effect of high order motion feature than the angle measurement accuracy. Radar station layout optimization combined with radial velocity augmentation measurement can effectively obtain high precision estimation of high order motion feature of target.

The method of tracking filter of ballistic target is summarized. The key points， advantages and disadvantages of the extended Kalman filter （EKF）， conversion measurement Kalman filter （CMKF）， ballistic EKF （BEKF）， and ballistic unscented Kalman filter （BUKF） are analyzed. The efficacy of the four filtering methods is further verified using simulated ballistic data. The filtering precision， tracking performance of different filters for reentry target are compared and the effect of ballistic coefficient and measure error on filter performance is analyzed. The method of forming covariance of measure error is proposed， which improves adaptability and precision of filter. Several problems that should be considered when choosing and designing target tracking filter methods are proposed to provide reference and bases for the selection and desing of radar filtering methods.