Optimal Land Grading Based on Genetic Algorithms

Abstract

Design of a successful and efficient surface irrigation system involves land grading as its first step. Land grading is also an effective method for efficient surface drainage. In this research, a nonlinear optimization model based on genetic algorithms is developed for land grading design of irregular fields. The model can be used to obtain plane as well as curved surfaces. In contrast to the available models, the proposed optimization model employs an objective function in which the total volume of earth in cut is minimized. The proposed formulation eliminates the need for parallel shifting of an optimal graded surface by trial and error to adjust the cut-to-fill ratio to a desired value. This is achieved by imposing a constraint on the cut-to-fill ratio. By eliminating the need for such parallel shifting, the optimality of the graded surface is kept undisturbed. Constraints may also be imposed on the shape of the graded surface, slopes of the surface in either direction, elevation of the surface at turnout, and at any other points. The developed optimization model is applied for grading an irregular field, which appeared in several previous publications concerning land grading design. The results indicate efficiency and robustness of the developed model.