Research On Optimization Of Land Use Structure In Yangzhou City Based On Multi-objective Genetic Algorithm Evolution

With the accelerated process of industrialization and urbanization,Human demand for land resources has grown from a single economic demand,to economic,social,ecological and other aspects of demand.Factors to consider in the process of rational allocation of land resources among different land use sectors are also increasing.Multl-objective and multi-constraint has become the main direction of land use structure optimization.The second land survey was carried out in the whole country in 2009.Satellite remote sensing technique is general used in this process,and the professional level of investigators has improved compared with previous years.The precision and accuracy of the data the second land survey is higher than the original data.Due to the change of land data's precision and statistical caliber after the second land survey,land use structure adjustment program in current land use master plan(2006-2020)has lost its currency.Urgently need to re-formulate land use structure adjustment program.Traditional land use structure optimization methods are difficult to solve multi-objective and multi-constraint optimization problem.A fast elitist non-dominated sorting genetic algorithm for multi-objective optimization is a search heuristic algorithm to solve optimization problems in the field of computer science and artificial intelligence,it is suitable for multi-objective and multi constrained optimization problems.This paper uses the system theory,sustainable development theory and ecological economic theory.On the basis of understanding and clarifying the structure and level of land use system,theoretical analyzes the multi-objectives of land use,then constructing the multi-objective system of land use structure optimization,and quantifying the multi-objective system into the objective functions or constraint equations in the optimization model layer-by-layer,so as to establish the optimization model of land use structure.According to the NSGA-? source code provided by Kanpur Genetic Algorithms Laboratory,combined with the research questions of this paper,writing algorithm program by C++ language,running program to solve optimization model of land use structure,there are a series of Pareto solution sets.Using TOPSIS method based on entropy weight to choose the best structure by calculating the close degree of each altermative and the ideal scheme.An empirical study on the optimization of land use str,ucture taking Yangzhou city as an example.Firstly,analyze the current situation of land use and the problems in land use in Yangzhou city.Then,constructing the optimization model of land use structure which coordinating the economic,social and ecological goals.Using genetic algorithm to solve the model,15 land use structure schemes are obtained.Using TOPSIS method based on entropy weight to calculate the closeness degree between the 15 schemes and the ideal scheme.The closeness degree of scheme 7 is the maximum,so it is the best land use structure.Research results show that:1)the accounting of the social and ecological benefits of land use is an important prerequisite for constructing the optimization model of land use structure which coordinating the economic,social and ecological objectives and achieving the goal of social benefit and ecological benefit of land use.2)NSGA-? is suitable for solving multi-objective and multi-constrained land use structure optimization problems.3)The land use structure scheme chosen by TOPSIS based on entropy weight meet the requirements of economic,social and ecological benefits.So this method can be used as a reference for decision maker to compile and revise land use planning.