Robust Estimation Model for Correlated Observations Based on Standardized Residuals

This paper builds robust estimation model for correlated observations structured by standardized residuals from the relationship between the observation errors and the function of difference solutions. By calculating and analyzing GPS network with simulative gross error. this model not only has good robustinity to both observation space and design space.but also can eliminate the infulence that the unsuitable prior standard error of unit weight takes on the robust estimation effect.