Tracking of a maneuvering target and multiple site track correlation

James Alan Roecker, Purdue University


When radar systems were first put into use, targets were tracked by the human operator on the radar screen. Now that radar systems are used in many military and civilian applications to track a large number of targets at once, there is a need for algorithms to automatically track the targets for the operator. In civilian operations (airport air traffic control) the air traffic controllers need a system to track the airplanes, estimate their course, and possibly detect future collisions. The military also needs this type of application along with the prediction of the location of a target for a fire control system. Estimation and prediction filters such as the Kalman filter can provide a good estimate of the target's state and predict the target's next state only when the target's motion is properly modeled. This study examines a method for modeling circular turns in a linear fashion for a Kalman filter implementation. This study also examines the use of multiple radar sites. Two different scenarios are addressed namely overlapping radar coverage and non-overlapping coverage. When multiple radar sites have overlapping coverage and are tracking a common target, a tracking system can provide a better estimate of the target's state than when only one site is used. Two methods are presented and compared for fusing the capabilities of multiple radar sites. When the radar sites do not have overlapping coverage target tracks will be broken during the period when the target is not in a scanning area. An algorithm is presented to connect broken tracks in such an environment. Simulation results for all algorithms are presented.




McGillem, Purdue University.

Subject Area

Electrical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server