元胞自动机具有能模拟复杂动态系统的强大能力,本文采用了多约束条件的元胞自动机模型,以广东佛山市2000年、2006年和2012年建设用地的变化为例,从自然、社会经济发展等方面综合考虑选取了高程、坡度、人口密度、道路交通、水系等对城市建设用地发展变化起决定作用的诸多因子,利用马尔科夫概率矩阵计算2000年~2006年建设用地变化,推算建设用地转移总量。结合Logistic-CA模型和决策树-CA模型,预测模拟了2012年的建设用地分布并与实际相比较,分析其整体精度和误差来源。结果显示基于CA模型的建设用地动态发展模拟具有良好的效果,可以为城市的发展规划,过程演变提供虚拟的实验手段和科学依据。
Cellular automata has a strong ability to simulate the complex dynamic systems,taking the changes of construction land in 2000,2006 and 2012 of Foshan city in Guangdong province as an example,this paper adopted the urban cellular automata model with many restriction for further research.Considering the nature,social and economic development and other aspects comprehensively,many factors including elevation,slope,population density,road traffic and drainage are selected and played as decisive roles in determining the development of the urban construction land.In order to reckon the total transfer of the construction land,the changes of the construction land from 2000 to 2006 are calculated by Markov probability matrix.Combined Logistic-CA model with Decision tree CA model,this paper also predicted and simulated the distribution of construction land in 2012 and compared with the actual situation,analysed its overall accuracy and error sources.The result shows that the dynamic change simulation of the construction land has good effect.It could provide virtual experiment means and scientific basis for the development planning and evolution of the city.