Parameters estimation of horizontal multilayer earth based on Kriging model updating method.

• A Kriging model updating method combined of Kriging surrogate model and genetic algorithm is proposed to estimate parameters for horizontal multilayer earth model. • The efficiency and accuracy of the proposed method are analyzed. • The optimization accuracy influenced by initial data samples is discussed. • Verification cases have been illustrated for the feasibility of final estimated earth models. A Kriging model updating method combined with Kriging surrogate model and genetic algorithm is proposed in this paper, helping to improve the computational efficiency of parameters estimation for horizontal multilayer earth model around electrical substation. Based on the theory of widely used Wenner configuration method for apparent resistivity measurement, the optimization model for horizontal multilayer earth model parameters is built up. To avoid time-consuming calculation, Kriging surrogate model is introduced to replace the complicated optimization model using small amount of data samples. Then genetic algorithm is adopted to update this Kriging model and estimate the earth model parameters until the criterion achieved. The accuracy and efficiency of the proposed method are analyzed. And several cases of multilayer earth model are studied furtherly and compared with traditional genetic algorithm to verify its feasibility. [ABSTRACT FROM AUTHOR]