基于人工神经网络的喀斯特地区水资源安全评价  

Assessment of Water Resources Security in Karst Area Based on Artificial Neural Network

作　　者：刘丽颖[1,2] 官冬杰[3] 杨清伟[1] 苏维词[4,5] 

机构地区：[1]重庆交通大学河海学院,重庆400074 [2]重庆工商大学数学与统计学院,重庆400067 [3]重庆交通大学建筑与城市规划学院,重庆400074 [4]重庆师范大学地理与旅游学院,重庆400047 [5]贵州科学院山地资源研究所,贵州贵阳550001

出　　处：《水土保持通报》2017年第2期207-214,共8页Bulletin of Soil and Water Conservation

基　　金：国家自然科学基金项目"喀斯特城市边缘带土地利用/覆盖变化及其环境效应"(41261038);"三峡库区生态安全后续发展动态模拟及其可视化预警评价"(41201546);国家科技支撑计划项目(2014BAB03B01);重庆市自然科学基金项目(cstc2012jjA20010);重庆工商大学2015年校级科研项目(670101577)

摘　　要：[目的]对喀斯特地区水资源安全状况进行评价,为喀斯特地区水环境管理与可持续发展提供科学依据。[方法]以喀斯特典型分布区贵州省为例,建立喀斯特地区水资源安全评价指标体系,构建BP人工神经网络模型,对贵州省9个州市水资源安全进行评价,并与熵权物元模型相比较。[结果]贵州省9个州市的水资源安全,有2个处于比较安全状态,6个处于一般状态,1个处于比较不安全状态。评价结果与熵权物元模型评价结果基本一致。[结论]该体系评价结果合理,评价方法简便直观,该模型对类似地区水资源安全评价有一定的参考价值。[Objective] Water resource security was evaluated to provide scientific basis tor water environment management and sustainable development in Karst area. [Methods] A case study of Guizhou Province in karst was carried out and an evaluation index system of water resources security was established. The BP artificial neural network model was constructed to evaluate the water resources securities of 9 cities in Guizhou Province, and the results were compared with entropy weight matter element model. [Results] Water resource securities of 9 cities were evaluated in Guizhou Province, among which two cities were assessed at relatively safe level, six cities at general level, and one city at relatively unsafe level. The evaluation results were basically the same to the results of the entropy weighted matter element model. [Conclusion] The results are reasonable and the method is simple and intuitive. The model has certain reference value for similar areas of water resources security evaluation.

关 键 词：水资源 安全 神经网络 喀斯特地区 模型构建 

分 类 号：TV213[水利工程—水文学及水资源] P642.25[天文地球—工程地质学]