Tracking Algorithms For Single Target Based On UKF

In this paper , three problems are studied: One is nonlinear filtering, another is how to build the models of target, the other is virtual application under different sensors.Because algorithm of filtering is one of core in the tracking system, the problem of generalizing the Kalman filter paradigm for nonlinear applications is considered. For the well-known limitations of the extended Kalman filter(EKF), recently a new nonlinear estimator is developed, that is Unscented KF(UKF), It yields performance equivalent to the EKF for linear systems, yet generalizes elegantly to nonlinear systems without the linearization steps required by the EKF. We show analytically that the expected performance of the UKF . We demonstrate the performance benefits in an example application of tracking a ballistic object in the reentry phase , and we argue that the ease of implementation and more accurate estimation features of the UKF recommend its use over the EKF in virtually all applications.For the passive sensor only obtains angle measurements, it is not steady for tracking an object in Cartesian Coordinate System . So, we study the questions of modeling and tracking target in Spherical Cordinate System.Finally, the problem of tracking the maneuvering target is considered. The algorithm is the IMM, with PDAFs used as building block for each of the models. We select the UKF as the filter of IMM/PDAF, Through tracking the "S" typically maneuvering target, we demonstrate the performance of the algorithm of IMM/PDAF with two passive sensors.