Parameters estimation of horizontal multilayer soilsusing a heuristic algorithm

Abstract

The grounding grid is buried in horizontal multilayer soils, so accurate parameters (resistivity , thickness, the number of soil layers) estimation of the soil is the basis for accurately calculating the grounding parameters and design of grounding grids. At present, the parameters estimation is usually solved by non-heuristic algorithms and heuristic algorithms. However, existing non-heuristic algorithms depend too much on the initial value and conventional heuristic algorithms could deviate from the correct solution too far, which would easily fall into local optimal solution, causing calculation errors. To solve the above problem, this paper first studies the relationship between the apparent soil resistivity and soil parameters based on the Wenner's method. Then, an innovative PSO-DE (Particle Swarm Optimization-Differential Evolution) optimization algorithm and its initialization method are proposed. The results of PSO-DE, conventional heuristic intelligent optimization algorithms, CDEGS, and literatures are compared, which verifies the superiority and accuracy of the PSO-DE. Therefore, the contribution of this work can lay a solid theoretical foundation for grounding parameters calculation.