Design of tracking system using bayesian position prediction for highly maneuverable aerial target

Abstract

Recently, the use of drones in the private sector has been rapidly increasing, such as toy drones and aerial photography drones. Such indiscriminate increases are causing many problems, such as shooting security facilities, invading privacy, flying in prohibited areas, and threatening people. In addition, terrorist groups such as ISIS, Hamas and Hezbollah have drones or already have core technologies, so the danger of terrorism using drones has also become a big problem. This research is to design a tracking system to block illegal drones. It is assumed that the flight target used in this study has highly maneuverable characteristics through a control input that the tracker does not know. In order to track a highly maneuverable target, it is necessary to estimate the future position through the current estimated target flight state. And we proposed an algorithm that can track highly maneuverable target by applying the existing guidance law of impact angle control using prediction information. We applied the weighting function using the variance information estimated by Bayesian method. The performance of the developed tracking system was verified by 6 DOF flight simulation. Finally, we implemented it through flight experiment and compared with simulation results.