The article aims to present a hybrid approach combining information entropy and an evolutionary algorithm to optimize a geodetic network's measurement structure to determine an engineering object's horizontal displacements. The objective function was defined, which in the case under consideration was the information entropy of the geodetic observation system in terms of the parameter vector's entropy with the true values. The optimal number of observations in the geodetic network depended on the observation system's increase in information. During the research, it was noticed that the application of the hybrid approach allowed the selection of only those observations with the highest information content. It shortened the measurement time without reducing the accuracy of the displacements obtained. The obtained results of numerical analyses showed the proposed solution's effectiveness for optimizing the geodetic network structure.