CMSC 2018 Presentation - A New Model for Spherical Instrument Measurement Uncertainty

Publication Details

Abstract

Monte Carlo uncertainty for spherical instruments (e.g. tracking interferometers, theodolites) typically uses a spherical-to-cartesian model whose point clouds change from spheres at the equator to disks as uncertainty point clouds approach zenith and nadir.  This does not comport with reality since the field-of-view and uncertainty though an instrument should be the same at all zenith angles.  A new uncertainty model is proposed that distinguishes between the on line-of-sight (LOS) and off LOS rays.  The new model treats the off-LOS rays so they result in spherical uncertainty clouds at all zenith angles.  Comparison of the spherical-to-cartesian uncertainty model and new uncertainty model for point-to-point and point-to-object geometries show that they both will construct the geometry identically within the measurement space.  The standard deviation for the geometry fitting, however, can be different for the two models, with generally the standard deviation of the new model being larger than standard spherical-to-cartesian model.