BP神经网络模型在采水地面沉降中的应用研究  

Application of BP neural network model to ground subsidence of mining water area

作　　者：周复旦[1,2] 赵长胜[1] 高卫东[1] 

机构地区：[1]徐州师范大学测绘学院,江苏徐州221116 [2]徐州市勘察测绘研究院,江苏徐州221002

出　　处：《测绘科学》2011年第6期233-234,240,共3页Science of Surveying and Mapping

基　　金：徐州师范大学2010年度研究生科研创新计划重点项目(2010YLA007)

摘　　要：随着我国各项建设对水资源的需求越来越大,导致由地下水开采而引起的沉降问题成为当前研究的热点课题。本文对某矿区采水地面沉降进行了模型设计,通过对部分实测数据的训练,优选出该模型的网络结构和网络参数,并且用Matlab软件编程实现对其他监测点的计算和预测。通过研究表明本文所建立的BP神经网络模型能较准确反映采水地面沉降的规律,同时也能较准确地预测地下水开采引起的地面沉降。With the growing water demand in China domestic construction, subsidence caused by groundwater over-exploitation has become a topic of current research. This article made model design in light of water area＇ s subsidence. By training some monitoring points＇ measured data, it selected the network structure and parameters of this model. By using the models, it calculated and predicted other monitoring points with the help of Matlab software. Research results indicated that the BP neural network model could accurately reflect the spatial law of water area＇ s subsidence and predict the subsidence caused by underground mining precisely.

关 键 词：BP网络模型 MATLAB软件 地面沉降 模型预测 

分 类 号：P631.6[天文地球—地质矿产勘探]