Abstract: In industrial measurement, the coordinate measurement error of the traditional moving station measurement will accumulate constantly with the increase of number of stations, which leads to the uncontrollability of the measurement accuracy. Based on the original observations and the coordinate observations respectively, this paper proposes two methods of data fusion of the overall adjustment in which the point precision is both higher and more stable than the traditional moving station measurement. Experiments show that the fusion results won't be completely equivalent within different fusion methods even if the mathematical model and stochastic model are equivalent. In order to get more reliable accuracy of coordinates of unknown points, we should choose different fusion methods based on the specific situation in practical application.