Small Target Tracking Algorithm Based On Auxiliary Knowledge Study

The application of traditional tracking method is restricted by the weak power from the targets and the strong power from the uninteresting targets which make the tracking of the small targets difficult. On the one hand, the advanced flyers such as stealth fighter and cruise missile and so on whose radar echoes are much weaker, are widely used in the modern war, on the other hand, the using of the target which moves in the low attitude lead to the surveillance area much complexer with stronger power of the clutter. To improve the tracking of the small targets by radar more effective, more timely, more accurate, the method of the Knowledge-Aided Tracking by using prior information to increase the difference between target and clutter, is studied by the researchers at home and aboard.This dissertation focus on the problem of small target tracking in the complex environment, by studing the prior information of the environment and the targets feature to aid the small target tracking, which mainly includes the working as follows:1. Aiming at the problem that the unstable tracking performance of the small target in the non-homogeneous environment on the different landform, the method that the classification information of the surveillance area attained from the Digital Terrain Elevation Data, the false alarm probability and so on, aid the tracking of the small target is presented, which can reduce the number of the false tracks heavily when compared with the traditional tracking methods.2. Aiming at the problem that it’s so difficult to make a distinction between the target’s measurement and the false measurement in the data association when there are so many spurious measurements, the method that target’s feature information aid the tracking is proposed, which can improve the right association probability between the target’s track and measurement, compared with the traditional data association algorithm which use the kinetics data only.3. Aiming at the problem that the tracking of the ground target is difficult for the strong power of the clutter, the high maneuverability and so on, the method that the small target tracking aided by the road map information is put forward. It can improve the accuracy of the track in the small target’s tracking, when compared to the traditional method which fix the number and the structure of the motion models along the whole surveillance area in the algorithm such as interacting multiple model.The effectiveness of the method that knowledge aided tracking of small target are validated by simulations. The results show that the tracking performance of the small target can be improved obviously by using the prior information of the environment and the feature information of the target in the tracking, compared with the traditional methods.