Elman神经网络在矿井突水水源判别中的应用

发布时间：2018-06-10 17:28

本文选题：安全工程 + Elman神经网络　；参考：《安全与环境学报》2017年04期

【摘要】：矿井突水是矿建与生产过程中最具威胁的自然灾害之一,准确判别突水水源是防治水害的关键。选取6种离子的质量浓度作为突水水源的判别因素,将河南省焦作矿区不同水层的39组水化数据以2种样本设计方案进行Elman神经网络模型的构建与检验。以不同的35组水源样品作为训练样本,运用Matlab软件进行Elman神经网络训练,将所建立的判别模型应用于(相应的)4组待测样本的判别,并与DDA、FDA、Bayes三种判别方法的判别结果进行分析比较。2种方案应用结果表明:将具有非线性动态特征的Elman神经网络应用于突水水源判别,在结合相应的水文地质条件前提下,可以准确判断突水来源;矿井多年的开采促使地下各水层水质呈动态变化,Elman神经网络判别模型能够反映这种变化特性,对探寻地下水运移与演化具有一定的应用价值。
[Abstract]:Mine water inrush is one of the most threatening natural disasters in the process of mine construction and production. The mass concentration of 6 kinds of ions was selected as the discriminant factor of water inrush water source, 39 groups of hydration data of different water layers in Jiaozuo mining area of Henan Province were constructed and tested by two different sample design schemes for Elman neural network model. Using 35 different water source samples as training samples and using Matlab software to train Elman neural network, the established discriminant model was applied to the discrimination of 4 groups of samples to be tested. Compared with the discriminant results of DDA-FDA-Bayes three discriminant methods, the application results of the two schemes show that Elman neural network with nonlinear dynamic characteristics is applied to the judgment of water inrush source under the premise of corresponding hydrogeological conditions. It can accurately judge the source of water inrush, and the mining of mine for many years makes the water quality of underground water layer dynamic change. Elman neural network discriminant model can reflect this kind of change characteristic, and has certain application value in exploring groundwater migration and evolution.
【作者单位】：河南工程学院安全工程学院;武汉理工大学资源与环境工程学院;中南大学资源与安全工程学院;
【基金】：国家自然科学基金项目(51604091) “矿灾害预防与控制河南省高校重点实验室培育基地”建设经费资助项目(200925) 河南省高等学校重点科研项目(16A440001,18A440010)
【分类号】：TD745.2

【相似文献】