A.A. Nikitina, P.A. Gritsyk

Most of trackers and the radar information secondary treatment are based on the following algorithms: the alpha-beta, the alpha-beta-gamma, and the Kalman filters. The last one is the recursive algorithm to solve the state system estimation problem for which the motion model is well-known. For simply connected systems Kalman filters are steady and have good convergence. Due to big computational capabilities of modern computers, Kalman filters find more applications in modern radar systems. At the same time there are many factors leading to divergence of the Kalman filter and, consequently, poor quality of secondary treatment in real situations. The paper contains the algorithm ob movement features estimation of the ballistic target, which maneuvers on the some path section. Movement features are the state vector of such object: its location and velocity on the predetermined local rectangular coordinate system. This algorithm of the radar information secondary treatment is based on using the classic Kalman filter and the maneuver detector which determines the moment of maneuver beginning and estimates the control influence value. At the maneuver detection moment the model is changed to another one, which takes into account the control acceleration. The results of simulation show high performance of this algorithm from response time on a maneuver point of view and the accuracy of getting estimates of the target movement features. In addition, the algorithm synthesized is optimal one by the minimum mean-square deviation criteria and can be applied in engineering systems.