遺伝的アルゴリズムを用いた都市の土地利用パターンの形成

概要

論文の詳細を見る
In the field of construction and urban planning, arrangement problems are essential and important, and many methods for them have been studied. However, especially that of urban planning are complicated. If there are many numbers of elements, it may be difficult to discover optimum arrangement in realistic time by former methods. And worse still, the complexity of cities themselves makes systemization very difficult. In recent years, applying Genetic Algorithm (GA), which is a new method imitating the process of the living-thing evolution, to arrangement problems is studied. In this study, the authors try to reproduce urban land use by using GA with two local evaluation functions, i.e., geographical location conditions and degrees of mutual combination. First, a city arrangement model is proposed, and the coding methods of genes are shown. Then the system is applied to the land use of Kobe City. The applicability of this system is proved.
1997-05-30