Location Analysis Of The Site Investigation Using Kriging

In the field of geotechnical engineering, the study of geological data provides an important basis for foundation design, foundation treatment, foundation reinforcement and so on. In general, in situ tests and laboratory tests will be performed in the construction areas, so as to get the geological data. But the geological data are discrete. Soil is formed by the sediments, which is derived from the weathering of the rock surface. According to the process, the parameters of the soil have properties of spatial correlation and variability. Kriging method can calculate the spatial correlation by using the known data. And it can also predict values of the unknown region according to the known data and the spatial correlation. Thus, the continuous curved surface of data can be obtained. This paper predicts geological information of no drilling areas by using Kriging method. If the new geological exploration points are added on the original geological information, different new geological information is obtained when new locations of the exploration points are different. Thus, there are optimal new locations of the exploration points. In other words, more detailed geological information can be obtained by using the same amount of geological exploration points.The soil which is formed in the natural process has strong anisotropy. At present, spatial interpolation method is mainly applicable to isotropic soil. Kriging method can be used to predict the isotropic soil parameters distribution by using the known geological information. For the soil profile, the spatial variabilities of horizontal direction and vertical direction are obviously different. Therefore isotropic Kriging method can not be used to predict soil parameters distribution of the unknown region. It needs to develop a Kriging method which can be used in anisotropic soil. And this paper focuses on studying anisotropic Kriging method.At first, according to the basic principle of Ordinary Kriging method and equivalent elliptic synthesis method, we study the method of anisotropic Ordinary Kriging. Moreover, the programs are written based on researches and the correctness of the programs is verified. Secondly, when exploration data of the estimated site is insufficient, the Ordinary Kriging method can not be used to predict the values of unknown area. But when the adjacent sites(or similar sites) have enough exploration data, the Bayesian Kriging method can be used to estimate the values of unknown area. Therefore, this paper develops the method of anisotropic Bayesian Kriging and the programs of this method have been written based on the basic principle of Bayesian Kriging method and equivalent elliptic synthesis method. In this paper, we use two known, stable and similar two-dimensional random field to verify the correctness of the programs. The proposed anisotropic Bayesian Kriging programs can be used to predict parameter distribution of the soil profile. Finally, this paper does a research on optimizing locations of new exploration points based on Ordinary Kriging method and Bayesian Kriging method combined with actual engineering test data, and optimization criteria has been established. Research results show that the new exploration point should be chosen at the location where the standard deviation is maximum. Optimizing new exploration location can obtain more geological information with less exploration points, and it will be in favor of saving total cost of the project construction. Therefore, the researches of this paper have great influence on the practical engineering.