ABSTRACT
Contraposing the problem that the observation data with the gross error and the performance weakly in the least-square estimation, two kinds of weight functions of the robust estimation are used to deal with the data. The robust estimation is a method of reducing or eliminating the effects of gross errors involved in the observation data on the unknowns by selecting the appropriate weight function. Through the examples, two weighting functions of Huber and Hampel are used to show that the robust estimation can improve the accuracy of the observation data by adjusting the influence on weighting and reducing the gross error to the adjustment values.
