Application of ellipsoidal approximation into target tracking for multi UAVs formation cooperative detection

Abstract

This paper studies the ground target positioning and tracking algorithm for cooperative detection of multi unmanned aerial vehicles (UAVs) formation, and a real time and rapid algorithm is discussed based on UAV airborne electro-optical sensors. Firstly in case of known probabilistic distribution on noise, unscented Kalman filter algorithm is applied to estimate the unknown state for target tracking process. Secondly to relax the strict condition on white noise in Kalman filtering theory, the target tracking or state estimation are reduced to derive the inner and outer ellipsoidal approximations for the state in case of unknown but bounded noise. More general cases are studied with the number of the ellipsoids, and some alternative forms are derived to obtain the approximate outer and inner ellipsoidal approximations. Finally, one simulation example confirms our theoretical results.