北京大学学报(自然科学版)

基于MATLAB实现的ANN方法在地下水质评价中的应用

罗定贵1,2, 王学军1, 郭青3   

  1. 1北京大学环境学院,北京,100871;2东华理工学院,土木与环境工程系,抚州,344000;3东华理工学院,电子与自动化系,抚州,344000
  • 收稿日期:2003-02-27 出版日期:2004-03-20 发布日期:2004-03-20

The Application of ANN Realized by MATLAB to Underground Water Quality Assessment

LUO Dinggui1, 2, WANG Xuejun1, GUO Qing3   

  1. 1College of Environmental Sciences, Peking University, Beijing, 100871; 2Civil and Environmental Engineering Department, East-China Institute of Technology, Fuzhou, 344000; 3Department of electronics and automation, East-China Institute of Technology, Fuzhou, 344000
  • Received:2003-02-27 Online:2004-03-20 Published:2004-03-20

摘要: MATLAB 6.5工具箱提供了径向基网络的实现函数,该算法具有自适应确定网络结构和无需人为确定网络初始权值的特点。将其应用于抚州市地下水环境质量评价,并尝试利用MATLAB的PREMNMX函数进行原始数据预处理、利用RAND函数在水质评价标准等级间内插构造足够数量的训练样本、检测样本及其目标输出、确立水质评价等级界限,取得良好的评价结果,对提高水质评价的精度与客观性具有十分积极意义。

关键词: 地下水, 环境质量评价, RBF网, 人工神经网络

Abstract: The realizing function of RBF network is provided in the toolbox of MATLAB 6.5 and the method of calculating this function possesses properties such as adaptation for determining the construction network and independence of initial weight value on person. A favorable outcome appeared after we apply this function to evaluating the quality of the underground water environment in FuZhou City, attempting to use the RAND function in MATLAB to construct enough training samples, checking samples and outputs of their targets through interpolation between grades of the water quality evaluation standard, use the PREMNMX function to pretreat the original data, determine the limits of water quality grades. The methods in this paper is meaningful in improving the precision and objectivity of underground water environment quality evaluation.

Key words: underground water, environment quality assessment, RBF network, artificial neural network

中图分类号: