基于神经网络与元胞自动机的城市扩展模拟

Urban Expansion Simulation Based on Artificial Neural Network and Cellular Automata

为了探究城市扩展的规律,为城市的规划做出前瞻性的预测,将神经网络与元胞自动机相结合,从不同时相遥感数据中挖掘城市扩展土地利用演变的规律,自动找到土地利用元胞的转换规则,并以该规则反演和预测城市的扩展演变。应用该方法对义乌市的扩展作了实证分析和模拟预测,与同期义乌城市发展状况基本相吻合。

In order to explore the rules of urban expansion and to serve urban planning, this paper researched on the combination of cellular automata with artificial neural network, and then set up the ANN-CA mode. The mode was applied to excavate the rule of the evolution of land use from different temporal remote sensing image and to simulate urban expansion. In this paper, the application example was Yiwu city. The result of simulate was considerably coincident with simultaneous expansion in research area in Yiwu.