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

运用人工神经网络方法预报表层岩溶地下水动态
引用本文:邱士利,夏日元,陈宏峰,姚昕,金新峰.运用人工神经网络方法预报表层岩溶地下水动态[J].广西师范大学学报(自然科学版),2005,23(4):99-102.
作者姓名:邱士利  夏日元  陈宏峰  姚昕  金新峰
作者单位:中国地质科学院,岩溶地质研究所,广西,桂林,541004
基金项目:国家自然科学基金资助项目(40272068);国土资源部攻关项目(20010303);国家科技攻关项目(2002BA901A13)
摘    要:将人工神经网络的理论与方法应用到表层岩溶带地下水动态预报中,从不同角度建立了2个预报BP网络模型,A模型刻画水位与降雨量、蒸发量和泉流量之间的关系,相对误差为±2%;B模型刻画水位自身的变化规律,相对误差为±5%,均可用于预报水位.A,B模型可为表层岩溶水研究及开发利用提供依据.

关 键 词:人工神经网络  BP网络模型  地下水动态  预报模型  表层岩溶带
文章编号:1001-6600(2005)04-0099-04
收稿时间:2005-04-07
修稿时间:2005-04-07

Prediction of Epikarst Groundwater Level Variation with Artificial Neural Network
QIU Shi-li,XIA Ri-yuan,CHEN Hong-feng,YAO Xin,JIN Xin-feng.Prediction of Epikarst Groundwater Level Variation with Artificial Neural Network[J].Journal of Guangxi Normal University(Natural Science Edition),2005,23(4):99-102.
Authors:QIU Shi-li  XIA Ri-yuan  CHEN Hong-feng  YAO Xin  JIN Xin-feng
Abstract:The principle and the method of artificial neural network have been applied to the prediction of variation of groundwater levels in epikarst zone.Two prediction BP network models,i.e.A and B,have been built from different points of view.Model A depicts the relations between water level and the precipitation,evaporation and spring flux,and the relative errors of its forecasting results are less than 2 percent.Model B depicts the variation regularity of water level itself,and the relative errors are less than 5 percent.Two models can be both used to predict the water levels prediction,which can provide an evidence for the development and(utilization) of water in epikarst zone,and water level datum for research on epikarst zone.
Keywords:ANN  BP model  variation of groundwater level  prediction model  epikarst zone
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