Improving the best filters which are estimating position and kinematics of maneuvering targets and design of a new approach, template filter

The most important part of the maneuvering air target tracking systems is the filtering part. Filters find the position and kinematics (velocity and acceleration) of the targets using the noisy mesurements taken from sensor ( radar ). The better positions and kinematics are calculated, the better route is found and target is determined. This is also important for the continuity of tracking. Among known constant gain filters, Alpha-Beta and Alpha-Beta-Gamma filters, among the adaptive filters, Kalman Filter, and multiple model, IMM (Interacting Multiple Model) Filter are analysed in deep. Their working principles, advantages and disadvantages, success and fail conditions are presented in theory. Their usage in real world scenarios and the way to choose parameters are shown. An improvement suggestion that brings adaptivity to Alpha-Beta and Alpha-Beta-Gamma filters is proposed. The improved filters, called Adaptive Alfa-Beta and Adaptive Alpha-Beta- Gamma filters, are analysed with their advantages and disadvantages and their performances are compared with those of basic Alpha-Beta and Alpha-Beta-Gamma Filters. Additionally a modification to the IMM filter which removes the interaction part of IMM filter, which is called Non-interacting Multiple Model (NIMM) filter, is proposed and compared with basic IMM filter. Finally a new filter, called Template Filter, is introduced. This filter?s performance is tested with real world scenarios and compared with those of all previous filters. All the filters are tested and compared with most difficult scenarios for trackers and air tracking radar works with real world conditions. Keywords: Target tracking, maneuver, Kalman, IMM, Alpha-Beta, Alpha-Beta-Gamma, NIMM, Template filter