北京城市增长边界预测

为了遏制北京”摊大饼”式的城市发展模式,促进土地资源的集约、高效利用,引导国土空间合理布局,划定北京城市增长边界成为重要举措。本研究从城市内生发展动力角度,将遗传神经网络嵌入CA模型,建立北京城市增长边界预测模型,选定了自然、人口经济、区位、邻域、土地利用类型和政策规划6类因素共18个影响因子,预测北京市城市增长弹性边界。同时,从土地自身生态承载能力角度,用建设用地适宜性评价方法,选定了地形、地貌、公园水域、土地利用现状、自然保护区、城市用地距离为影响因子,划定北京市城市增长刚性边界。结果表明,利用该模型预测北京市2025年城市增长边界,总的面积匹配值为96%,模型精度较高;2025年北京市弹性增长边界总面积为1738.98 km²,刚性增长边界总面积为3297.01 km²。基于GANN-CA模型的城市增长边界划定方法对确定城市未来扩张方向有指导作用,可为城市规划和发展政策的制定提供依据。

The Prediction of Beijing Urban Growth Boundary

Urban growth boundary (UGB) is an effective tool to control urban sprawl. China has carried out the UGB delineation for 14 cities since 2014. UGB includes rigid boundary and elastic boundary. The rigid boundary is an external constraint line and the elastic boundary is the developing line of the interior. In this study, from the angle of the urban endogenous development motivation, we selected 6 types of influence factors such as nature, population economy, location, neighborhood, land use type and policy planning, totally 18 factors. We tried to use genetic algorithms, CA and BP neural network to establish the UGB model to define Beijing elastic boundary. From the angle of ecological carrying capacity of land, we selected terrain, topography, parks and water, land use status, urban land distance and nature reserves as the influence factors and we tried to use construction land suitability evaluation method to define Beijing rigid boundary. The results show that by using our urban growth boundary model to predict Beijing elastic urban growth boundary, the percent area match was 96%, kappa value is 0.812, the UGB Model accuracy is good and the prediction area of Beijing elastic UGB is 1738.98 square kilometers. Through the suitability evaluation, the rigid boundary area is 3297.01 square kilometers.