This paper presents an Extended Kalman Filter for tracking a
maneuvering target, where the kinematics of a typical target aircraft maneuver have been incorporated into the filter state equations. Such a formulation allows the target's motion to be accurately determined through estimation of heading and lateral acceleration. This is an improvement over the the typical approach of modeling target motion with acceleration terms represented by random processes,
such as that used in the Singer model. In the following pages,
a three-dimensional target maneuver model is formulated in conjunction with the kinematic equations of a sensor tracking a target in spherical coordinates. Three degree-of-freedom simulation results of the proposed filter, simplified for planar target maneuvers, are compared to a filter modeling target motion with the Singer model.