The Research On Maneuvering Target Tracking Algorithm

Target tracking technology is an indispensable technology in land, sea, air traffic control system. Maneuvering target tracking has been one of the hot topics in the field of target tracking. With the increasing mobility of all kinds of targets, effective target tracking becomes more and more difficult. The research on maneuvering target tracking algorithm is a challenging task both in theory or in practice. The main research content of this paper has four parts.Part 1, it explains the basic principles of target tracking and introduces severl basic motion model, such as CV model, CA model, Singer model, CS model, CT model, and summarizes the characteristics of each model. Then study the filtering algorithm, including Kalman Filter, Extented Kalman Filter (EKF), Unscented Kalman Filter (UKF). UKF has better filtering accuracy than EKF through simulation.Part 2, it studys deeply on Interacting Multiple Model (IMM), including algorithm steps, filter design and other issues. UKF has been adopted In the use of IMM algorithm for maneuvering target tracking. simulating the IMM algorithm and comparing the result to single model algorithm, the result proves the IMM algorithm has better tracking accuracy.Part 3, it improves IMM-CV/CT algorithm aiming at high speed maneuvering target, by choosinging the best Turning model from a discreted turning model set, and then interect with CV model. Finally, the tracking results are obtained by UKF filtering. simulating the improved algorithm, it proved that it has a better tracting effect in tracting high speed maneuvering target, tracting accuracy is high improved.Part 4, aiming at the Improved IMM algorithm, through Software programming it realizes the processing of radar target data on the Visual Studio 2010 platform, and give the tracking effect chart.