Iterative multi-level magnetic matching for UAV navigation

A geomagnetism aided inertial navigation system for UAVs is studied and emphasis is laid on the research of magnetic matching method. In view of specialty of UAV and error properties of inertial navigation system, a simplification is given to the whole one-dimension matching process, which leads to an affine transformation with a compensation step. An iterative multi-level matching method is proposed to get the transform parameters, in which contour constraint and two-stage search strategy are used to reduce the computation burden. In compensation part, a nearest neighbor rule is applied to correlate vehicle position and geomagnetic measurements. Global searching ability and localization precision are both considered. Good performances are observed in both simulation and real experiments.