Data Screening for Kinematic Orbit Determination of LEOs

摘要

The orbit determination (OD) accuracy of kinematic and reduced-dynamic method based on the spaceborne GPS data is strongly dependent on the quality and continuity of GPS observations, and the screening of the Low-Earth-Orbiter (LEO) GPS observations is substantially important. In this paper, an elaborate pre-screening procedure of zero-differenced spaceborne GPS observations was developed. This new efficient data-screening algorithm consists of two steps: in the first step, the "majority voting" method is used to locate the gross errors in zero-differenced GPS data; in the second step, the observation marked as gross error is corrected using the "Quasi-Accurate Detection of gross errors (QUAD)" method. In this step, an iterative least-square adjustment procedure is used to obtain residuals which are screened to detect bad observations thoroughly. For illustration of the feasibility and efficiency of the new algorithm, the processing of a numerical example of CHAMP satellite is presented in this paper.