Improved bearings-only target tracking with iterated Gaussian mixture measurements

Abstract

In this study, a novel iterated Gaussian mixture measurements filter is proposed to represent the measurement likelihood function (LF) with Gaussian mixtures more precisely. The proposed approach recalculates the range interval using updated track components, and the LF is remodelled with Gaussian mixtures for the new range interval. Then, every regenerated measurement component is used to update all of the predicted track components for generating new updated track components. Experimental results demonstrate that the proposed method outperforms existing methods in bearings-only target tracking.