土地利用主体微观行为的群集智能模拟及其宏观涌现机制研究

The contradiction between supply and demand of land resources becomes more and more sharpening and the regional man-land relationship is more tensional in china．In order to delve deeply into the interaction mechanism between groups activity and natural process, this proposal integrates GIS, Cellular Automaton and Multi-agents to develop the swarm intelligence model for land use action. Then the model is used to simulate the micro-decision behaviors of heterogeneous land users, and the spatial adaptation process between man and land. Furthermore, the simulation process reveals the spatial interaction evolution rule of geographic environment and interactive behaviors of groups, and explores the internal mechanism by which the local micro-interaction of land users emerges into macro-geographic spatial effects in one given region. Therefore, the purpose is to establish the spatial correlation mechanism between land micro-users' behavior and land use macro-planning in order to contribute to the spatial administrable order of land development, and improvement of land use planning theory and methods. The research includes: (1) To analyze the spatial behavior determinants of land use agent, and to construct behavior indicators, (2) To analyze the correlation of the interaction between land use agents, (3) To build multi-source data fusion model and database of land use, (4) To design the intelligence of land use agent using case-based reasoning, (5) To construct swarm intelligence model of micro-macro-emergence behavior of land use agent, (6) To test the spatial effect of the swarm intelligence model, (7) To exploit the emergence mechanism and regulations of land use agents' behavior.

我国正面临的形势是土地资源供需矛盾日益尖锐、人地关系日趋紧张。为深究群体活动与自然过程相互作用机理，本项目通过GIS、元胞自动机、多智能体等技术集成构建土地利用行为群集智能模型，模拟异质性土地利用主体微观决策行为及其人地关系相互适应的空间过程，揭示不同群体之间交互行为与地理环境相互影响的空间演化规律，探索一定地域土地利用主体微观局部行为涌现成宏观地理空间效应的内在机理，旨在建立微观主体土地利用行为与宏观层面土地利用规划的空间关联机制，为规范国土空间开发秩序、完善土地利用规划理论与方法上有所贡献。研究内容包括：(1)土地利用主体的空间行为决策因子分析及行为指标构建；(2)土地利用主体之间的互动行为关联性分析；(3)土地利用多源数据融合与数据库构建；(4)基于案例推理的土地利用主体智能化设计；(5)微-宏观涌现行为的群集智能模型构建；(6)模型的空间效应检验；(7)土地利用行为规范的提炼。

将物联网技术引入到空间数据仓库系统的构建之中，对物联网空间数据仓库系统的数据源、数据处理层、数据存储层及应用分析层进行详细的设计，构建物联网空间数据仓库系统框架体系；在GIS平台上，通过元胞的自然条件、邻域关系、交通通达性及控制因素来产生元胞的转换规则，构建元胞自动机（CA，Cellular Automata）模型,对研究区域的城市用地变化进行模拟与预测，同时，探讨了元胞自动机“生态位”适宜度模型的转换规则，评估低丘缓坡建设用地的适宜性。运用智能体(Agent)模拟了耕地保护行为；运用GIS进行上海市绿地公共空间可获得性分析；基于GIS进行了深圳市创造性工作者的空间分析；将GIS空间分析技术、层次分析法（AHP）及特尔菲法（Delphi）相结合，分别构建宜耕未利用地开发适宜性评价模型、新增耕地指数及补充耕地潜力模型、宜耕未利用地开发潜力分区评价模型，对云南山区宜耕未利用地开发适宜性与潜力分区进行了分析；基于GIS的滇中地区耕地时空变化驱动力；征地拆迁冲突主体行为研究；基于GIS的城市群主体功能区土地利用结构与空间管治的研究。