基于蚁群算法的可扩展多目标土地利用优化配置

发布时间：2018-04-03 20:14

本文选题：土地资源　切入点：土地利用优化配置　出处：《浙江大学》2017年硕士论文

【摘要】：土地资源是人类生产活动最基本的资源。在城市化进程中,人口增长、不合理的土地利用和低土地利用率对紧张的人地关系提出了巨大的挑战。在土地资源有限的情况下,只有科学分析当前的土地资源分布状况,合理配置有限的土地资源,才能实现土地资源的节约集约利用,实现城市可持续发展。土地利用优化配置是一个复杂的多目标问题。在进行土地利用优化配置时,需要综合考虑经济、社会和生态等多个因素,每个因素以及评价这个因素的指标均可视为一种目标。而经济、社会和生态评价的复杂性,是土地利用优化配置多目标的复杂性的根本原因。由于土地资源存在空间异质性,现有的大多数优化配置模型侧重于土地利用的数量结构优化,在空间布局优化方面还处于探索阶段。本文在国内外土地利用优化配置理论的指导下,结合蚁群算法,建立了一个基于蚁群算法的可扩展多目标土地利用优化配置模型,并以杭州市萧山区为研究区域,获取土地利用优化配置方案,对比分析了不同多目标体系下、优化前后的土地利用结构和空间布局,主要研究成果如下:(1)可扩展多目标体系的构建。介绍了土地适宜性评价、空间集聚和最小转换成本的函数模型,在此基础上构建基本的多目标优化体系,然后对常见的目标进行抽象建模,并定义不同类型扩展目标的目标函数,最后构建成可扩展的多目标体系。(2)应用蚁群算法解决土地利用优化配置多目标问题的处理。对蚁群算法解决土地利用优化配置问题进行详细分析。详细分析蚁群算法的重要参数,主要包括蚁群规模、信息素挥发因子、信息启发因子和期望启发因子等,并对蚁群算法的核心函数根据多目标优化要求进行处理,将多目标优化问题转为单目标问题求解。最后根据蚁群算法的流程,构建完成基于蚁群算法的可扩展多目标土地利用优化配置模型。(3)在模型的基础上,以杭州市萧山区为研究区域,进行实例验证。在对萧山区土地利用现状进行分析的基础上,以2015年为基期年,在不同的多目标优化体系下,对萧山区土地利用进行优化配置,得到2020年萧山区土地利用优化配置方案,并将不同的优化配置方案进行对比分析,验证模型的有效性和灵活性。
[Abstract]:Land resources are the most basic resources of human production activities.In the process of urbanization, population growth, irrational land use and low land use rate pose a great challenge to the tense human-land relations.In the case of limited land resources, only scientific analysis of the current distribution of land resources, reasonable allocation of limited land resources, land resources can realize intensive utilization of land resources, and realize sustainable development of cities.Land use optimization is a complex multi-objective problem.Economic, social and ecological factors need to be taken into account in the optimal allocation of land use. Each factor and the index to evaluate this factor can be regarded as a goal.The complexity of economic, social and ecological evaluation is the fundamental reason for the complexity of land use optimization and allocation.Due to the spatial heterogeneity of land resources, most of the existing optimal allocation models focus on the optimization of the quantitative structure of land use and are still in the exploration stage in the aspect of spatial layout optimization.Under the guidance of land use optimal allocation theory at home and abroad and combined with ant colony algorithm, this paper establishes a scalable multi-objective land use optimal allocation model based on ant colony algorithm, and takes Xiaoshan District of Hangzhou as the research area.The optimal land use allocation scheme is obtained, and the land use structure and spatial layout before and after optimization are compared and analyzed under different multi-objective systems. The main research results are as follows: 1) the construction of extensible multi-objective system.This paper introduces the function model of land suitability evaluation, spatial agglomeration and minimum conversion cost. Based on this, the basic multi-objective optimization system is constructed, and then the common objectives are abstractly modeled.The objective function of different types of extended objects is defined, and an extensible multi-objective system. 2) the application of ant colony algorithm to solve the multi-objective problem of land use optimization is presented.The ant colony algorithm to solve the land use optimization allocation problem is analyzed in detail.The important parameters of ant colony algorithm are analyzed in detail, including ant colony size, pheromone volatilization factor, information heuristic factor and expectation heuristic factor, and the core function of ant colony algorithm is processed according to multi-objective optimization requirements.The multi-objective optimization problem is transformed into a single-objective problem.Finally, according to the flow of ant colony algorithm, a scalable multi-objective land use optimal allocation model based on ant colony algorithm is constructed. On the basis of the model, a case study is carried out in Xiaoshan District of Hangzhou City.Based on the analysis of the present situation of land use in Xiaoshan District, taking 2015 as the base period, and under different multi-objective optimization systems, the optimal allocation of land use in Xiaoshan District in 2020 is obtained.The effectiveness and flexibility of the model are verified by comparing and analyzing different optimal configuration schemes.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F301.2;TP18

【相似文献】