基于模糊神经网络的海水水质综合评价  

Integrated assessment on sea water quality based on the fuzzy neural network

作　　者：刁美娜[1,2] 温小虎[1] 刘有刚[3,4] 王德[1] 高猛[1] 

机构地区：[1]中国科学院烟台海岸带研究所,山东烟台264003 [2]中国科学院研究生院,北京100049 [3]国家海洋局 [4]青岛环海海洋工程勘察研究院,山东青岛266033

出　　处：《海洋通报》2012年第2期228-232,共5页Marine Science Bulletin

基　　金：中国科学院知识创新项目(KZCX2-YW-224-04);国家自然科学基金(41001013)

摘　　要：根据俚岛镇海湾22个海洋观测点的水质监测数据,依据人工神经网络与模糊理论,建立模糊神经网络模型,对研究区海水水质进行了综合评价。结果表明,区域内40,91%的采样点属于Ⅰ类海水水质标准,其余为Ⅱ类海水水质。其中,水质Ⅱ类区位于临络湾与俚岛湾的半封闭海域,因该区造船业对水体造成污染且海水对流强度较弱,不利于污染物的扩散;而海水水质Ⅰ类区域靠近黄海,工业活动较少,水体对流较强,利于污染物扩散。In this study, with 22 water quality data from sampling sites of Lidaozhen Bay, the quality of sea water based on fuzzy neural network was assessed. The results show that 40.91% of the samples belong to class I water quality standard and others belong to class Ⅱ. The class Ⅱ region lies in the semi-closed sea in the Linluo Bay and the Lidao Bay. In these areas, pollutants from shipbuilding industry and the weak convection of water are the main reasons for the pollution. But in water quality class I region, there are less industrial activities and stronger water convection, which are all beneficial to the dispersion of pollutants.

关 键 词：模糊神经网络 临络湾 俚岛湾 海水水质评价 

分 类 号：X824[环境科学与工程—环境工程]