地理学报 ›› 2010, Vol. 65 ›› Issue (6): 695-707.doi: 10.11821/xb201006007

• 土地利用 • 上一篇    下一篇

基于多智能体的居住区位空间选择模型

刘小平, 黎夏, 陈逸敏, 刘涛, 李少英   

  1. 中山大学地理科学与规划学院,广州510275
  • 收稿日期:2009-04-22 修回日期:2010-03-24 出版日期:2010-06-25 发布日期:2010-06-25
  • 作者简介:刘小平(1978-), 男, 湖南邵阳人, 副教授, 中国地理学会会员(S110007282M)。研究方向为地理模拟与空间智能, 已发表论文50 多篇, 其中SCI收录20 多篇。E-mail: yiernanh@163.com
  • 基金资助:

    国家自然科学基金项目(40901187); 国家自然科学基金重点项目(40830532); 广东省自然科学基金项目(9451027501002471)

Agent-based Model of Residential Location

LIU Xiao-ping, LI Xia, CHEN Yi-min, LIU Tao, LI Shao-ying   

  1. School of Geography and Planning, Sun Yat-sen University, Guangzhou 510275, China
  • Received:2009-04-22 Revised:2010-03-24 Online:2010-06-25 Published:2010-06-25
  • Supported by:

    National Natural Science Foundation of China, No.40901187; Key National Natural Science Foundation of China, No.40830532; Guangdong Provincial Natural Science Foundation, No.9451027501002471

摘要:

多智能体系统(Multi-Agent system, MAS) 是一种进行复杂系统分析与模拟的强有力工具,尤其在社会科学领域得到了广泛的应用。本文提出了基于多智能体的居住区位选择模型(Agent-Based Model of Residential Location—ABMRL),将多智能体建模的方法应用于居民居住区位决策行为和地价动态变化的研究中,旨在探索与模拟居民在居住选择过程中的复杂空间决策行为,以及居民之间、居民与地理环境的相互作用而导致城市居住空间分异的演化过程。ABMRL模型由表征各类居民的多智能体层和表征地理环境的元胞自动层组成,对应人地关系中的两个基本要素——人类与自然环境。该模型认为居民迁居的动力源于内部的经济社会压力和外部的居住环境刺激。利用ABMRL模型模拟和验证了居住空间分异、圈层城市空间结构、城市绅士化等经典城市理论,并以广州市海珠区为实验区,模拟了该区域居民居住空间分异的演化过程和地价的动态变化。

关键词: 多智能体, 复杂系统, 城市, 区位选择, 居住分异, 广州

Abstract:

Multi-agent system (MAS) is a powerful tool capable of analyzing and simulating complex systems, and has been extensively applied in the regime of social sciences. In this paper we present an agent-based model of residential location (ABMRL) and apply this model to study the dynamic changes of residential location and land price, aiming to explore and simulate the complicated spatial decision-making behaviors in residential location as well as the evolution process of urban residential segregation, which is resulted from interactions among residents and between residents and geographical environment. The ABMRL model consists of agent layer representing various classes of residents and cell automata layer representing geographical environment, which correspond to the two basic elements in man-earth relationship, i.e., human being and natural environment. In this model, psychological concepts were introduced for study of the behaviors of residential location, as it is generally considered that residential relocation is facilitated by internal social and economic pressure and external residential environment. ABMRL model was used to simulate and validate a series of classic urban theories such as residential segregation, concentric urban space structure, gentrification, etc., and to simulate the evolution of residential segregation and dynamic changes in land price in Haizhu District of Guangzhou City, which was taken as a test example for the study.

Key words: multi-agent, complex system, urban area, residential location, residential segregation, Guangzhou