基于T-S模糊神经网络的民勤地下水水质综合评价  

Comprehensive assessment of groundwater quality in Minqin Basin based on T-S Fuzzy Neural Network

作　　者：汪新波[1] 粟晓玲[1] 

机构地区：[1]西北农林科技大学水利与建筑工程学院,陕西杨凌712100

出　　处：《干旱地区农业研究》2013年第1期188-192,198,共6页Agricultural Research in the Arid Areas

基　　金：国家自然科学基金项目(50879071);西北农林科技大学基本科研业务费科技创新重点项目(QN201168)

摘　　要：为了摸清石羊河流域民勤盆地近25年来的地下水水质变化状况,为当地水土资源合理开发和生态环境保护提供决策参考依据。将T-S模糊神经网络应用于民勤盆地1983、1990、2000年及2008年的地下水水质评价中,并与支持向量机(SVM)模型的评价结果进行检验比较。结果表明:民勤盆地地下水水质总体较差,并且盆地南部地区水质整体优于北部,除红崖山水库附近地区,80%以上区域水质为Ⅴ类水;141、147、156、168号井等在山区边缘的部分站点水质随时间有改善趋势。两种模型评价结果基本一致,但T-S模糊神经网络收敛速度更快,可以有效应用于地下水水质综合评价。In order to find out the variation of groundwater quality of Minqin basin in Shiyang River Valley in recent 25 years and to provide the decision-making reference for rational exploitation of local water resources and eco-environ- mental protection, T - S Fuzzy Neural Network model was applied to the comprehensive assessment of groundwater quality in the year of 1983, 1990, 2000 and 2008, and the Support Vector Machines （SVM） model was applied to test the re- suits. The results showed that the overall groundwater quality of Minqin Basin was poor and it was overall better in the southern region than in the northern region. Except for the area surrounding Hongyashan reservoir, the groundwater quali- ty of more than 80% regions was poorly achieved grade V. The groundwater quality of partial wells at the edge of deserts, such as No. 141, 147, 156 and 168, showed a improving trend. The results of the two models were generally concordant, but the T- S Fuzzy Neural Network model exhibited a fast convergence, therefore it can be effectively ap- plied to the comprehensive assessment of groundwater quality.

关 键 词：T-S模糊神经网络 支持向量机 地下水水质评价 民勤盆地 

分 类 号：P641.8[天文地球—地质矿产勘探]