Robust-Biased Estimation Based On Quasi-Accurate Detection

Abstract

In order to combat the influences of both outlier and multicollinearity on geodetic adjustments, a new robust-biased estimation method is proposed by combining outlier identification with biased estimation. The estimation scheme is roughly divided into two steps. First, quasi-accurate detection of gross error (QUAD) is used to detect outliers and correct observations. Then the “clean” observations and biased estimations are used to obtain more accurate estimates of unknown parameters. Several selection schemes of the biased parameters included in the biased estimators based on QUAD are given in detail. A numerical example illustrates that the new robust-biased estimation method not only can resist the bad influence of outlier and effectively overcome the difficulty caused by multicollinearity simultaneously, but also is far more accurate than least-squares estimation, biased estimation, robust estimation, and generalized shrunken type-robust estimation.