基于风险因子层次分析法的生态环境需水量模糊神经网络模型  

FNN model of eco-environmental water demands based on AHP

作　　者：胡德秀[1] 周孝德[1] 米艳芳[1] 

机构地区：[1]西安理工大学水利水电学院,陕西西安710048

出　　处：《西安建筑科技大学学报（自然科学版）》2011年第2期224-230,共7页Journal of Xi'an University of Architecture & Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(50779051);陕西省教育厅专项科研计划项目(07JK354);西安理工大学科学研究基金项目(106-210509)

摘　　要：针对以往多因子分析中存在因重复计算所致的冗余和因孤立计算所致的结果不合理,提出和建立基于风险因子层次分析法的生态环境需水量模糊神经网络模型;采用层次分析法分析生态环境需水量风险因子间的相互作用,建立其量化指标组合权重关系,并将权重值作为所建FNN模型的初始权值输入,从而有效消除了系统随机赋予初始权值对FNN模型结果的影响,并以黄河上游龙羊峡河段为例,进行模型拟合与风险分析.结果表明,所建模型合理可行,其拟合精度较高,具有较强的工程实用价值.Aimed at the redundancy caused by double counting and the irrationality caused by isolation in the past analysis of multi-factors,in this article,an FNN（fuzzy neural network） model of eco-environmental water demands based on AHP（analytical hierarchy process） under cascade development was put forward.AHP was used to analyse the interaction of risk factors of eco-environmental water demands,to establish quantitative indicators＇ compound weight on various factors,and,the weight sets were input as initial weight values of the impact factors of FNN model,thus effectively eliminating the impact randomly assigned initial values on the model results,And to the upper reaches of the Yellow River,with Longyangxia as an example,the comparative analysis of model fitting and risk studies were carried out.Project instance application shows that the model built is reasonable and feasible,and has a better fitting accuracy and some practical value for engineering.

关 键 词：生态环境需水量 层次分析法 模糊神经网络 风险分析 梯级开发 

分 类 号：X171.4[环境科学与工程—环境科学]