首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于GRNN网络模型的土壤重金属空间分布的研究
引用本文:胡大伟,卞新民,李思米,许泉.基于GRNN网络模型的土壤重金属空间分布的研究[J].土壤通报,2007,38(2):334-340.
作者姓名:胡大伟  卞新民  李思米  许泉
作者单位:南京农业大学,农学院,江苏,南京,210095
基金项目:江苏省自然科学基金;江苏省生态环境安全研究项目
摘    要:以江苏省南通市为研究区,利用采样点实测数据,借助GRNN神经网络模型并结合3S技术对农田土壤重金属的空间动态分布进行了深入研究。结果表明,GRNN神经网络模型能够智能地学习各个采样点的空间位置与该点各重金属含量之间的映射关系,并能够稳健地对各个空间插值点处的土壤重金属含量进行预测;结果显示南通市农田土壤重金属污染总体较轻,但也存在局部地区的严重污染。在运用GRNN神经网络模型进行空间插值了解重金属空间动态分布的基础上,可以根据污染的状况确定农产品的生产布局和规划。

关 键 词:GRNN神经网络模型  3S技术  土壤重金属
文章编号:0564-3945(2007)02-0334-07
修稿时间:2006年3月27日

Spatial Distribution of Farmland Heavy Metals Based on GRNN-ANN Modeling
HU Da-wei,BIAN Xin-min,LI Si-mi,XU Quan.Spatial Distribution of Farmland Heavy Metals Based on GRNN-ANN Modeling[J].Chinese Journal of Soil Science,2007,38(2):334-340.
Authors:HU Da-wei  BIAN Xin-min  LI Si-mi  XU Quan
Abstract:It is essential to study farmland soil heavy metals in order to provide information about scientific planning of farmland.Hence,the objective is to describe spatial dynamic distribution of farmland soil heavy metals.We selected Nantong,Jiangsu province as the research region to carry out the study by using the GRNN Artificial Neural Networks Modeling and 3S technology.The results showed that the GRNN modeling not only can provide the mapping relationship between spatial position and heavy metal content in the sampling sits intelligently,but also can predict heavy metals content in every spatial interpolating point robustly.Based on the GRNN spatial interpolation and Arcgis analysis,our conclusion shows that,in general,the farmland soil in Nantong city belongs to the low pollution level,but some belong to the severe pollution level.
Keywords:GRNN neural networks modeling  3S technology  Soil heavy metals
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号