基于ANN-CA的胶州湾西岸港城土地利用情境模拟

Scenarios Simulation of Land Use in West Coast Port City of Jiaozhou Bay Based on ANN-CA

针对青岛市胶州湾西岸港城发展对土地利用的不同需求,采用人工神经网络(ANN)与元胞自动机(CA)的耦合模型,分别从城市导向与港口导向2种情境出发,对土地利用格局的进行模拟,分析不同情境下港城土地发展趋势与差异,揭示港城发展对土地功能演化及港城土地冲突的影响,在此基础上预测港城土地功能重叠区范围。结果表明青岛市胶州湾西岸地区土地发展格局较为鲜明,2015~2025年城市功能用地主要向西南方向扩展,港口功能用地主要向西北方向蔓延;港城土地功能重叠区主要分布在胶州湾西岸区域的西部、西北部与东南角。本研究成果可为港城土地资源规划利用提供建议。

Aiming at the demands of land use for the port city development on the west coast of Jiaozhou Bay in Qingdao City,we used the coupling model of artificial neural network(ANN)and cellular automata(CA)to analyze the characteristics of land evolution and expansion from the two guiding scenarios of city dominance and port dominance respectively.Then,we realized the differential simulation in the land evolution model.Finally,we put forward a method for predicting the overlapping area of land function in port city.The results show that from 2015 to 2025,the land development pattern in the west coast of Jiaozhou Bay in Qingdao City is more distinct,urban land mainly expands to the southwest,and port land mainly spreads to the northwest.The overlapping area of land function of port city mainly distributes in west,the northwest and southeast corner of the west coast area of Jiaozhou Bay,mostly in zonal distribution.The research results can provide reference for the formulation of different preferential policies for land use in port city.