Robust Estimation of Variance Components with Application in Global Positioning System Network Adjustment

Abstract

A robust strategy for controlling the influence of outliers on model parameters and variance components is proposed based on the principle of robust M estimation. The robust estimators for parameters and variance components are derived. A practical robust estimator similar to Helmert’s one for maximum likelihood estimation is obtained. The robustness and efficiency of the new robust estimation procedure is demonstrated and compared with those of the classical least squares method by applying a practical global positioning system network in China.