Performance Comparison of Maneuver Detection Algorithms

Abstract

This paper compares several algorithms for target maneuver detection. The algorithms are based on input estimation coupled with Kalman filter or IMM filtering. The performance of each method is analyzed computing a Pareto frontier, which is based on different quality criteria such as root mean square (rms) of state estimation error and time of maneuver detection. It is shown that computation of such Pareto frontiers allows to compare the performance of the different detectors in a systematic way.