摘要: 为了综合评价区域水资源承载力，采用投影寻踪聚类模型，借助加速遗传算法，优化投影方向，将多个影响因素在投影函数的作用下投影到一维数值，根据投影值的大小评价出方案的优劣。以青岛市为例，采用该模型，得出青岛市各区域的水资源承载力由高到低依次为：崂山区、市区、即墨市、莱西市、平度市、黄岛区、胶州市、城阳区；除崂山区和市区外，其余地区的承载力均趋于饱和状态；该市水资源对社会的支撑能力较强，但其保障程度较差。实验结果可以为改善青岛市严重缺水这一局面提供一些新视角，有利于帮助提出新的解决方案。
Abstract:
Through utilizing Projection Pursuit Classification model based on the Real Coding Based Accelerating Genetic Algorithm, the authors transform synthetically the multi-factor into only one index value which can comprehensively estimate the Water resource carrying capacity in line with optimum projection direction and function. Taking Qingdao as an example, the index values show the Water resource carrying capacity from up to low is Laoshan, urban districts, Jimo, Laixi, Pingdu, Huangdao, Jiaozhou, Chengyang; excepting the first two, Water resource carrying capacity of the rest of Qingdao all rush into saturation. Its water resource has an advantage in social supporting, while its guarantee extent stands disadvantage. The result can provide a new prospect for improving severe water shortage of Qingdao and helping to present some new solutions.