LAUSR

A method for improving asteroid orbits by gradual replacement of observations with their simulated values is considered. The method consists in successive rejection from processing of observations with maximal residuals and replacement of these observations with their values calculated in the last iteration of the process of orbit improvement. In addition, the traditional method of gradual rejection of observations with maximal residuals that does not preserve any information on these observations is considered. The accuracies of the methods are compared for a number of improvements of orbits of asteroids Nos. 115489 and 179699 from observations of their separate oppositions and pairs of oppositions. It is demonstrated that the method of replacement of observations with simulated values yields as a whole more exact results in comparison with the traditional methods of observation rejection.