基于多方法耦合的矿井突水水源识别方法研究  

Research on Identification Method of Mine Water Inrush Source based on Multi-method Coupling

作　　者：黄亚坤 李启兴 梁建斌 刘德旺[4] 夏向学[1] 张庆 任正瑞 康佳 吴威 徐保同 HUANG Yakun;LI Qixing;LIANG Jianbin;LIU Dewang;XIA Xiangxue;ZHANG Qing;REN Zhengrui;KANG Jia;WU Wei;XU Baotong(Hebei Key Laboratory of Mine Disaster Prevention and Control,North China University of Science and Technology,Langfang 065201,China;Huangyuchuan Coal Mine,Guoneng Yili Energy Co.,Ltd.,Erdos 017000,China;Jinjitan Coal Mine,Shaanxi Future Energy&Chemical Co.,Ltd.,Yulin 719099,China;Geological Resources Department,Shanxi Coking Coal Group Co.,Ltd.,Taiyuan 030000,China)

机构地区：[1]华北科技学院河北省矿井灾害防治重点实验室,河北廊坊065201 [2]国能亿利能源有限责任公司黄玉川煤矿,内蒙古鄂尔多斯017000 [3]陕西未来能源化工有限公司金鸡滩煤矿,陕西榆林719099 [4]山西焦煤集团有限责任公司地质资源部,山西太原030000

出　　处：《煤》2024年第11期6-10,共5页Coal

基　　金：中央高校基本科研业务费资助项目(3142022003,3142023055)。

摘　　要：为有效预防矿井突水事故发生,快速准确识别矿井突水水源是至关重要。以陕西某矿为例,选取K^(+)+Na^(+),Ca^(2+),Mg^(2+),Cl^(-),SO_(4)^(2-),HCO^(-)_(3)六种水化学成分指标作为矿井突水水源识别的样本变量。笔者采用主成分分析(PCA)结合Fisher判别分析法建立突水水源判别,利用不同的水化学特征样本中的27个数据为训练样本,对该方法进行检验和应用,并与传统的Fisher判别分析法的结果进行比较。研究结果表明:利用Fisher判别分析正确率为92.6%,而利用PCA与Fisher突水水源判别方法对样本数据进行判别,正确率达96.3%,并对10个预测样本数据进行判别,正确率100%,判别结果正确率较高,对煤矿突水水源识别具有指导意义。In order to prevent mine water inrush accident effectively,it is very important to identify mine water inrush source quickly and accurately.A mine in shaanxi province as an example,selecting K^(+)+Na^(+),Ca^(2+),Mg^(2+),Cl^(-),SO^(2-)_(4),HCO^(-)_(3)six water chemical composition index as mine water inrush water recognition sample variables.The author used principal component analysis(PCA)combined with Fisher discriminant analysis to establish the identification of water inrush source.27 data from different water chemical characteristic samples were used as training samples.The method was tested and applied,and the results were compared with the traditional Fisher discriminant analysis.The results show that the correct rate of Fisher discriminant analysis is 92.6%,and the correct rate of PCA and Fisher water inrush source discrimination method is 96.3%,and the correct rate of 10 predicted sample data is 100%,and the correct rate of discrimination results is high,which has guiding significance for the identification of coal mine water inrush source.

关 键 词：主成分分析(PCA) FISHER判别分析 矿井突水 水源判别 水化学特征 

分 类 号：TD745[矿业工程—矿井通风与安全]