High Maneuvering Target Tracking Using a Novel Hybrid Kalman Filter-fuzzy Logic Architecture

In this paper, a fast target maneuver detection technique and high accurate tracking scheme is proposed with the use of a new hybrid Kalman filter-fuzzy logic architecture. Due to the stressful environment of target tracking problem such as inaccurate detection and target maneuver, most of existing trackers do not represent desired performance in different situations. In practice, while the conventional Kalman filters (KF) perform well in tracking a target with constant velocity, their performance may be seriously degraded in the presence of maneuver. To reach an accurate target tracking system in such a stressful environment, fuzzy logic-based algorithms with intelligent adaptation capabilities have recently been issued. Although these methods yield reasonable performance in tracking maneuvering targets, their accuracy in non-maneuvering mode was not satisfactory. In this research, based on information about the target maneuver dynamics, a new hybrid tracker (HT) is introduced. The proposed algorithm combines two methodologies into one architecture synergistically. In other words, the KF is used when the target velocity is approximately constant, whereas fuzzy estimator is used when the target maneuvers. Simulation results show that the proposed method is superior to some conventional approaches in tracking accuracy.