地理学分析中基于自主体建模的原理讨论.PDF

王铮 中国科学院科技政策与管理科学研究所 地理学分析中基于自主体建模的 原理讨论 A DISCUSSION ON THE PRINCIPLE OF AGENT- BASED MODELING FOR GEOGRAPHICAL ANALYSIS 王铮 Zheng Wang Institute of Policy and Management, Chinese Academy of Sciences Difficult on geography analysis  Modern scientific methods from Newton  Humboldt hope to establish that kind of geography like Physics, so we have Physical Geography. Then, Physical Geography failed to develop a like physics has as a complete analysis system WHY? Friedrich von Humboldt， 1769－1859 地理学分析的困难  牛顿开始了现代科学分 析，他建立了Physics体 系  洪堡希望建立Physics那 样的地理学，但是, Physical Geography 不能 实现Physics的分析功能 这是为什么 Friedrich von Humboldt， 1769－1859 Difficult on geography analysis  Geography problems often do not have the background of continuous space, calculus analysis method based on continuous media is failure  Geography developed the analytical methods on geography map (ISARD). The map analysis suitable for the analysis of the space with not- continuous, but lacks the power of logical analysis .  Quantitative Geography is of based on mathematical statistics, it lack of the function of dynamical analysis 地理学分析的困难  地理学问题经常没有性质连续的空间背景， 基于连续背景的数值计算也往往是不可能 的；  Isard说地理学发展了特有的地图分析方法， 地图方法适合分析空间背景问题但是缺少 逻辑分析功能；  数量地理学基本上基于数理统计方法因此 缺乏动态分析功能 Agent-Based Modelling  19th century, the development of statistical physics from thermodynamics ，after Humboldt  A large number of particles system, the particle systems can be statistics  Agent concept evolved by particles, but the agent can be identified, there is