摘要:為了保證城鎮居民飲水安全,針對水質分級標準存在的模糊性,引入模糊神經網絡理論,建立水環境質量評價模型。選取吉林省白山市白云峰水庫為研究區,通過調查采樣選定6處監測點,將其模型評價結果與尼梅羅綜合指數法的評價結果進行對比分析。結果顯示,模糊神經網絡評價水環境質量是可行的,水質評價結果精度更準確,打破了傳統方法的局限性。模糊神經網絡模型具有較強的學習能力,能夠提高地下水水質評價的準確性,為日后保護和管理水環境提供了科學依據。
關鍵詞:模糊神經網絡;水環境質量評價;監測點
中圖分類號:TP18 文獻標識碼:A 文章編號:1009-3044(2014)20-4813-02
Application of Fuzzy Neural Network in Water Environmental Quality Assessment
ZHAO Xu1 ,CHEN Li-li2
(1.Geological and Mineral Resources of Liaoning Province Survey Institute, Shenyang 110031,China; 2.Heilongjiang Institute of Geological Survey, Harbin 150036,China)
Abstract: In order to ensure the safety of drinking water for urban residents, the fuzziness of classification of water quality standard, introduce the fuzzy neural network theory, establish the model of water environment quality evaluation. Selects the Jilin province Baishan City baiyunfeng reservoir as a study area, by sampling selected 6 monitoring points, the evaluation of the model evaluation results and the Nemero index analysis and comparison of results. The results showed that, fuzzy neural network evaluation of water environment quality is feasible, water quality evaluation result more accurate, to break the limitations of traditional methods. The model of fuzzy neural network has strong learning ability, can improve the accuracy of groundwater quality evaluation, provided the scientific basis for the protection and management of water environment.
Key words: fuzzy neural network; water quality evaluation; monitoring point
我國當前經濟社會的發展正處在城市化、工業化、現代化進程中,有效地保護和合理利用水資源,防止項目建設和生產造成的人為水資源破壞,最大限度地減少和降低對水環境的影響,保證工程項目的順利建設和安全運行,促進水資源的循環利用和生態環境的可持續維護,水環境質量科學準確的評價必不可少[1]。該文綜合考慮神經網絡的特點,把模糊理論引入評價模型中,以水質評價指標作為模型的輸入變量建立模糊神經網絡,以白山市白云峰水庫為研究區,評價其水環境質量。
1 模糊神經網絡
1974年,S.C.Lee以和E.T.Lee首次把模糊集和神經網絡聯系在一起; 1985年,J.M Keller和D.Huut提出把模糊隸屬函數和感知器算法相結合。自1992年開始,J.J.Backley發表了多篇關于混合模糊神經網絡的文章,它們也反映了人們近年來的興趣點。……