Abstract: Under the principle of genetic algorithm, by using the genetic algorithm toolbox, after proving the applicability of genetic algorithm on a numerical example of 3 zones, 3-section economy of nonlinear Lowry model, the authors use this model to perform a simulated experiment on 9 zones, 3-section economy under a fan-shaped urban framework. The results show that, at first, if the residential charm of every district is at the same degree, traffic factor will play a very important role in residential structure of urban area; secondly, the district at the best traffic location owns the highest population density. Because of the limitation of land and maximization of economic goal, this type of district is the first choice to the industry which have high added value. This article shows a good explanatory ability of nonlinear Lowry model and lays a good foundation for the practical simulated research of urban.

Key words: Lowry model, genetic algorithm (GA), NLP, land use

摘要： 利用遗传算法思想, 借助以MATLAB语言为基础的遗传算法工具箱, 在验证遗传算法在三区域、三部门模拟运算适用性之后, 用非线性Lowry模型对扇形城市空间结构下的九区域、三部门进行模拟实验。结果表明: 在各区域居住魅力相同的情况下, 交通因素对城市居住空间结构变化起到至关重要的作用; 交通区位最好的区域也是人口密度最大的区域, 因为用地的有限性和地区产值的最大化追求, 使这些区域也是高附加值产业首选之地。研究结果展示了非线性Lowry模型良好的解释能力, 为该模型推向具体城市模拟研究打下良好基础, 提高了其实用性。

关键词: Lowry模型, 遗传算法(GA), 非线性规划, 土地利用