Robust geodetic parameter estimation under least squares through weighting on the basis of the mean square error

Abstract

Four robust estimation techniques are evaluated in respect of their efficiency and effectivenes through the estimation of geodetic parameters within the framework of a geodetic network. A theoretical geodetic network in which gross errors have been simulated is adopted for the study. Using each of the robust estimation techniques in turns, two versions of the network contaminated with gross errors are adjusted and the results compared with the case without gross errors. The results indicate that the robust estimation techniques have been able to cope quite effectively with the gross errors introduced into the observations. The different levels of effectiveness for the individual techniques are also indicated.