Optimizing Land Use Allocation Based On Multi-objective Ant Colony Algorithm

The optimized allocation of land use is not only an important method to accelerate land intensive use and make land sustainable use come true, but also a hot point that the land science and the management currently faces. How to formulate a scientific and feasible planning and avoid falling into a specific digital game is the key issues to be solved. This paper will focus on the core problem-the optimal allocation of land use, by reviewing the existing model methods, most them focused on the number of land use structure optimization and lacked researches in the optimization allocation for spatial pattern of land use, which caused the planning achievements not to be properly carried out in the space.This paper will propose a new optimizing allocation model-ant colony algorithm, which combinates with GIS, will guarantee the coherence between the quantity and the space in certain degree.The paper focus around the study goal-the optimal allocation of land use, starts the relevant work by using ant colony algorithm, the main contents of the study include the following aspects:Firstly, The article introduces detailedly the land-use theory of optimal allocation of space research, technology research, the progress of the swarm intelligent optimization,ant colony algorithm and its main applications in different fields. By summarizing the documents home and abroad, the paper points out the existent problems in current researchs, then according to the meaning and characteristic, forms the research process of land use allocation. This section also introduces the basic theories and methods related to this article:the sustainable development thoery, the system theory, the ecology economic theory, landscape ecology, and the method of multi-objective optimization as well as the intelligence optimization method, the antcolony algorithm,etc.Secondly, the article proposed a multiobject model for land use optimization, the specific contents contain:introducing the basic ideas, basic principles for construction of the evaluation index system of land sustainable use, build an evaluation index frame of land sustainable use and the evaluation modle. This article also introduced the minimum cost planning model and the compactness model based on spatial clustering, according to the three models to construct system for multi-objective optimization. Bring forward the constraints system for space optimization (including the number of constraint system and conversion bindings of land-use type), and analying detaily the conversion of land use types and making the general rules of land-use conversion system. Finally, this article design the optimization integrated model of land use allocation, combinating the effectiveness of the traditional goals with the objectives of the space layout, the number of binding constraints and the joint effect of spatial pattern to the overall purpose of optimization.Thirdly, the article introduce detaily the design idea of the multiobjective ant colony optimization model. At first, the article introduce the core idea of basic ant colony and adaptive ant colony algorithm, then in order to adapt land use optimization allocation, we do some improvement aims to the limitation of the basic model, and design and explanation detaily for the improved multiobjective ACO, includes each parameter of the model, such as population size, heuristic factor, evaporation factor of pheromone, ect, and also contains how to deal with multiobject and the constraints. Subsequently, in order to meet the demand of land use data, the improved ACO is coupled with GIS to establish the spatial optimal allocation. This section design detaily the key link, including the coding framework of the ants, the initial formation of ants, the fitness evaluation function for each ant, heuristic information function, pheromone updating rule as well as the selection probability function and so on.Finally, this article carries out the application based on the land-use optimization model. According to the data of the general land use planning of Yicheng, Hubei province, selecting the representative town-Zhengji as an experimental zone. Firstly, the experiment evaluates the land use structure and spatial pattern. Based on the present land use tructure, according to the requirements of the experimental study of land use optimization model multi-objective ant colony optimization algorithm coupled with the GIS, studying the specific objective functions of the experimental zone, constraints, as well as the issue of data processing model for land use optimization model. Then, the article analyses the optimization results of the town, and compares the land sustainable use optimization with land utilization of status quo to verify the rationality of the classified index-based land sustainable use system and the adaption of the improved ACO optimization. At last, the aricle analyses the effection of different combination of preferences, and selects the best result based on the efficiency of algorithm, compares the result with GA optimization program and the basic ant colony algorithm to verify the advantage of improved multi-objective self-adapt ACO, which shows that the coupled model ACO and GIS is feasible in land use optimization allocation.