南通大学学报（自然科学版）

水资源短缺已经成为影响和制约我国部分地区社会和经济发展的主要因素,为了确定引起水资源短缺的主要风险因子,以达到控制风险的目的,以北京市1979—2009年水资源数据为例,应用BP神经网络良好的数据拟合功能,建立缺水量与导致水资源短缺各因子关系的BP神经网络模型.将各单个因子取值增加1%,然后通过该模型计算出缺水量的变化率,从而确定北京市水资源短缺的主要风险因子,依次为第三产业及生活用水、污水处理率、降水量等.该模型所得结果与现有方法的结论比较吻合,但计算更为简便.

Water shortage has been a major factor influencing and restricting the social and economic development in some areas.In order to understand the main risk factors of water shortage and to achieve the purpose of risk control,taking the Beijing municipal water resources data in 1979-2009 for example in this paper,the BP neural network model between the water-deficit and the each factor relation that causes water shortage are established,then the change rate of water-deficit by increasing 1% values of single factors via the newly-built model is calculated,the main risk factors for Beijing municipal water shortage are concluded as follwing tertiary industry,domestic water,treatment-rate of domestic sewage,and rainfall precipitation,etc.The results obtained by the model are in agreement with that by the existing approach,but the present calculation is simpler.