基于智能体的城市避灾绿地规划分析与建模研究
发布时间：2018-12-29 15:33
【摘要】：在几十年的高速城市化以后,中国的城市面临着越来越多的城市问题,规划设计者们开始反思传统城市规划方法的弊端,探讨提升城市规划技术的途径。计算机仿真模型是定量分析研究的有效技术手段,而智能体模型作为目前人工智能与社会学领域的研究热点,也受到越来越多的城市规划领域学者的关注。近年来中国自然地质灾害频发,城市绿地作为防灾避险场所的地位越来越重要,但与之相对应的规划技术支撑却十分薄弱。基于以上两点,本研究以智能体建模在避灾绿地规划分析中的应用为切入点,探讨智能体模型在城市规划定量分析中的应用。 研究通过对城市绿地系统规划、综合防灾规划等相关规划内容的整理,基于智能体模型构建理论,提出了构建城市避灾绿地规划分析模型的建模思路,在此基础上经过对比分析选择了Repast S作为智能体建模软件,继而构建了基于多智能体与GIS技术的仿真模型,对城市避灾绿地规划方案进行分析评价。通过不同尺度的规划案例数据在模型中的测试,验证了所构建的基于智能体的城市避灾绿地规划分析模型的运行与预期设想基本一致,也可为类似规划的分析提供有益思路。本文的主要研究成果如下： (1)探讨了智能体建模在城市规划中的应用特性、应用条件及可能的应用方向,明确了城市避灾绿地规划的定义及其具体编制内容与规划流程,指出了目前规划方法的不足与缺陷。在研究了智能体模型的构建原理的基础上结合避灾绿地规划的定量分析提出了避灾绿地规划分析模型的理论框架。 (2)研究了Repast S软件在城市规划中的应用、常用的CAD制图软件的数据导入方法及其与GIS的集成方法,实现了从常用制图数据到Repast S的应用数据转换、Agent在模型中的移动与算法选择、Agent的规则制订等关键技术。 (3)通过中观与宏观两个实际设计项目的数据,对已构建的智能体仿真模型进行了实证研究,进一步明确了从规划方案数据导入、到模型调试——输出结果的全部工作流程,提出了行之有效的城市避灾绿地规划分析方法,验证了智能体仿真模型在避灾绿地规划分析中的作用。
[Abstract]:After decades of rapid urbanization, Chinese cities are facing more and more urban problems. Planners begin to reflect on the drawbacks of traditional urban planning methods and explore ways to improve urban planning technology. Computer simulation model is an effective technique for quantitative analysis, and agent model is a hot topic in the field of artificial intelligence and sociology, which has attracted more and more scholars' attention in urban planning field. In recent years, with the frequent occurrence of natural geological disasters in China, urban green space is becoming more and more important as a place for disaster prevention and risk avoidance, but the corresponding planning technical support is very weak. Based on the above two points, this research focuses on the application of agent modeling in the planning and analysis of disaster avoidance greenbelt, and discusses the application of agent model in the quantitative analysis of urban planning. Based on the theory of agent model construction, this paper puts forward the idea of building urban green space planning analysis model based on the arrangement of urban green space system planning, comprehensive disaster prevention planning and other related planning contents. On the basis of comparative analysis, Repast S is chosen as the agent modeling software, and then the simulation model based on multi-agent and GIS technology is constructed to analyze and evaluate the planning scheme of urban disaster avoidance green space. Through the test of different scale planning case data in the model, it is verified that the operation of the model based on agent is basically consistent with the expected assumption, and it can also provide useful ideas for the analysis of similar planning. The main results of this paper are as follows: (1) the application characteristics, application conditions and possible application directions of agent modeling in urban planning are discussed. This paper clarifies the definition of urban green space planning for disaster avoidance, and points out the shortcomings and defects of the present planning methods. Based on the study of the construction principle of the agent model and the quantitative analysis of the disaster avoidance green space planning, the theoretical framework of the disaster avoidance green space planning analysis model is put forward. (2) this paper studies the application of Repast S software in urban planning, the data import method of CAD drawing software and its integration with GIS, and realizes the application data conversion from common drawing data to Repast S. The key technologies of Agent in the model, such as moving and algorithm selection, Agent rule making and so on. (3) through the data of the two actual design projects, the paper makes an empirical study on the agent simulation model, and further clarifies the whole work flow from the data import of the planning scheme to the debugging and output of the model. An effective method of urban green space planning and analysis is put forward, which verifies the role of agent simulation model in the planning and analysis of disaster avoidance green space.
【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TU984.116
本文编号：2395006
[Abstract]:After decades of rapid urbanization, Chinese cities are facing more and more urban problems. Planners begin to reflect on the drawbacks of traditional urban planning methods and explore ways to improve urban planning technology. Computer simulation model is an effective technique for quantitative analysis, and agent model is a hot topic in the field of artificial intelligence and sociology, which has attracted more and more scholars' attention in urban planning field. In recent years, with the frequent occurrence of natural geological disasters in China, urban green space is becoming more and more important as a place for disaster prevention and risk avoidance, but the corresponding planning technical support is very weak. Based on the above two points, this research focuses on the application of agent modeling in the planning and analysis of disaster avoidance greenbelt, and discusses the application of agent model in the quantitative analysis of urban planning. Based on the theory of agent model construction, this paper puts forward the idea of building urban green space planning analysis model based on the arrangement of urban green space system planning, comprehensive disaster prevention planning and other related planning contents. On the basis of comparative analysis, Repast S is chosen as the agent modeling software, and then the simulation model based on multi-agent and GIS technology is constructed to analyze and evaluate the planning scheme of urban disaster avoidance green space. Through the test of different scale planning case data in the model, it is verified that the operation of the model based on agent is basically consistent with the expected assumption, and it can also provide useful ideas for the analysis of similar planning. The main results of this paper are as follows: (1) the application characteristics, application conditions and possible application directions of agent modeling in urban planning are discussed. This paper clarifies the definition of urban green space planning for disaster avoidance, and points out the shortcomings and defects of the present planning methods. Based on the study of the construction principle of the agent model and the quantitative analysis of the disaster avoidance green space planning, the theoretical framework of the disaster avoidance green space planning analysis model is put forward. (2) this paper studies the application of Repast S software in urban planning, the data import method of CAD drawing software and its integration with GIS, and realizes the application data conversion from common drawing data to Repast S. The key technologies of Agent in the model, such as moving and algorithm selection, Agent rule making and so on. (3) through the data of the two actual design projects, the paper makes an empirical study on the agent simulation model, and further clarifies the whole work flow from the data import of the planning scheme to the debugging and output of the model. An effective method of urban green space planning and analysis is put forward, which verifies the role of agent simulation model in the planning and analysis of disaster avoidance green space.
【学位授予单位】：武汉大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TU984.116
【参考文献】
相关期刊论文 前10条
1 李晖;李志英;刘大邦;易娜;杨树华;李国彦;;基于元胞自动机的月亮山景区景观格局动态模拟[J];安徽农业科学;2011年03期
2 柳玲,艾及熙;GIS技术在城市规划领域的应用及发展[J];重庆建筑大学学报(社科版);2001年02期
3 于卓;吴志华;许华;;基于遗传算法的城市空间生长模型研究[J];城市规划;2008年05期
4 龙宏;王纪武;;基于空间途径的城市生态安全格局规划[J];城市规划学刊;2009年06期
5 徐礼鹏;刘启蒙;孙娇娇;;基于GIS的安庆市应急避难场所空间布局特征分析与优化[J];测绘与空间地理信息;2012年02期
6 朱宁;王大博;乔双;;粒子群优化算法在可供选择城市规划方案中的应用[J];东北师大学报(自然科学版);2010年02期
7 黎夏,叶嘉安;基于神经网络的元胞自动机及模拟复杂土地利用系统[J];地理研究;2005年01期
8 王波;张亮;孙霞;;CAD向GIS数据的转换入库[J];地理空间信息;2011年03期
9 宇林军;孙丹峰;彭仲仁;李红;;基于局部化转换规则的元胞自动机土地利用模型[J];地理研究;2013年04期
10 蒋云良,徐从富;智能Agent与多Agent系统的研究[J];计算机应用研究;2003年04期
相关博士学位论文 前2条
1 徐昔保;基于GIS与元胞自动机的城市土地利用动态演化模拟与优化研究[D];兰州大学;2007年
2 费文君;城市避震减灾绿地体系规划理论研究[D];南京林业大学;2010年
相关硕士学位论文 前1条
1 车平川;基于GIS的城市公园绿地布局优化研究[D];南京林业大学;2010年
，本文编号：2395006
本文链接：https://www.wllwen.com/guanlilunwen/chengjian/2395006.html
