地理学报 ›› 2011, Vol. 66 ›› Issue (8): 1033-1044.doi: 10.11821/xb201108003

• 城市研究 • 上一篇    下一篇

城市形态、交通能耗和环境影响集成的多智能体模型

龙瀛1,2, 毛其智1, 杨东峰3, 王静文4   

  1. 1. 清华大学建筑学院, 北京 100084;
    2. 北京市城市规划设计研究院, 北京 100045;
    3. 大连理工大学建筑系, 大连 116024;
    4. 北京林业大学园林学院, 北京 100083
  • 收稿日期:2010-04-27 修回日期:2010-07-28 出版日期:2011-08-20 发布日期:2011-09-22
  • 作者简介:龙瀛(1980-), 男, 博士, 高级工程师, 中国地理学会会员(S110007674M), 主要研究方向为规划支持系统和城市系统微观模拟。E-mail: longying1980@gmail.com
  • 基金资助:

    国家自然科学基金项目(51078213)

A Multi-agent Model for Urban Form, Transportation Energy Consumption and Environmental Impact Integrated Simulation

LONG Ying1,2, MAO Qizhi1, YANG Dongfeng3, WANG Jingwen4   

  1. 1. School of Architecture, Tsinghua University, Beijing 100084, China;
    2. Beijing Institute of City Planning, Beijing 100045, China;
    3. Department of Architecture, Dalian University of Technology, Dalian 116024, Liaoning, China;
    4. School of Landscape, Beijing Forestry University, Beijing 100083, China
  • Received:2010-04-27 Revised:2010-07-28 Online:2011-08-20 Published:2011-09-22
  • Supported by:

    National Natural Science Foundation of China, No.51078213

摘要: 城市能耗占全球能耗的比重随着城市化率的不断提高而增大,交通能耗作为城市能耗的重要构成部分,已有较多研究证明城市形态对其具有显著影响,这些研究多属于城市间层次,而少有城市内的研究对城市形态与交通能耗、环境影响的关系进行定量识别。本文拟建立城市形态、交通能耗和环境的集成模型,对单一城市内的不同空间组织(即城市形态),如土地使用方式、开发密度、就业中心的数量和分布等,对潜在的通勤交通能耗和环境影响的关系进行定量识别。该模型采用多智能体(multi-agent) 方法,一方面针对同一假想空间采用蒙特卡洛方法根据约束条件生成多个城市形态,并采用就业地斑块数目、平均斑块分形指数、香农多样性和平均近邻距离等14 个指标表征城市形态。另一方面,固定数量的居民agent 在所生成的每个城市形态内,选择居住区位和就业区位,根据通勤距离和社会经济特征选择交通方式,进而计算通勤交通能耗和环境影响,在城市层面统计通勤交通能耗和环境影响总和。最后分析城市形态与通勤交通能耗和环境影响的定量关系,主要得到以下结论,① 对于不同的城市空间布局和密度分布,通勤交通能耗的弹性范围约为3 倍;② 城市形态评价指数中,就业中心斑块的数量是对通勤交通能耗影响最大的变量;③ 多种城市形态所对应的通勤交通能耗基本呈正态分布。此外,还对城市形状对通勤交通能耗的影响进行了识别,并针对假想空间的多个典型城市形态(如紧凑与分散、单中心与多中心、TOD政策、绿隔政策),进行了通勤交通总量的计算,进而对典型规划理念进行了定量对比。本模型不仅可以用于识别城市形态与通勤交通能耗和环境影响的定量关系,定量对比典型的规划理念,还可以用于空间规划方案的能耗和环境影响评价。

关键词: 土地使用, 开发强度, 交通能耗, 环境影响, 多智能体, 蒙特卡洛

Abstract: Cities are consuming more energy with increasing urbanization process. The urban transportation energy is the primary part of urban energy consumption. Extensive researches found that it has strong relationship with the urban form, which fall into intra-cities level. However, little attention was paid to the relationship between urban form, transportation energy consumption, and environmental impact in the inner-city level. This paper aims to investigate the impact of urban form, namely land use pattern, development density distribution, on the residential commuting energy consumption (RCEC). We developed a multi-agent model for the urban form, transportation energy consumption and environment interaction simulation (FEE-MAS). Numerous urban forms with distinguished urban land use pattern and development density distribution are generated using the Monte Carlo approach in the hypothetical space. On one hand, the RCEC for each urban form is calculated using the proposed FEE-MAS. On the other hand, we select 14 indicators (e.g. Shape Index, Shannon's Diversity Index, Euclidean Nearest Neighbor Distance) to evaluate each generated urban form using FRAGSTATS, which is loosely coupled with the FEE-MAS model. Then, the quantitative relationship between the urban form and RCEC is identified based on 14 indicators and RCEC of each urban form. Several conclusions are drawn from simulations conducted in the hypothetical space. (1) RCEC may vary three times for the same space with various spatial layouts and density distribution. (2) Among selected 14 indicators for evaluating the urban form, the patch number of job parcels is the most significant variable influencing the RCEC. (3) The RCECs of all urban forms generated obey a normal distribution. (4) The shape of urban form also exerts influences on the RCEC. In addition, we evaluated several typical urban forms, e.g. compact/sprawl, single center/ multi-centers, traffic oriented development, green belt, in terms of the RCEC indicator using our proposed model to quantify those popular planning theories. The FEE-MAS model can also be applied for evaluating urban spatial alternatives in terms of energy consumption and environmental impact.

Key words: land use, development density, transportation energy consumption, environmental impact, multi-agent, Monte Carlo