基于Fisher判别法的岱庄矿涌(突)水水源识别

Source of Daizhuang Mine Water Inflow or Inrush Based on Fisher Discrimination

由于岱庄矿十下灰、十三灰和奥灰水源水质根据目前常规水质化验资料无明显的区别元素,容易产生误判。为实现对岱庄矿涌水水源的准确识别,在收集该矿地质及水文地质资料的基础上,首先对主要涌水含水层水化学特征进行分析,其次选取Na^(+)+K^(+),Ca^(2+),Mg^(2+),Cl^(-),SO_(4)^(2-),HCO_(3)^(-)这6种特征离子含量作为Fisher判别分析的因子,随后利用SPSS25.0软件建立岱庄矿涌水水源Fisher判别函数模型,最后选取该矿3组测试水样对模型进行验证。结果表明:利用Fisher判别分析理论建立的涌水水源识别模型,可准确识别这些水质相近的涌水水源,回代准确率达91.3%,3组测试水样判别结果也与工程实际情况一致,表明此模型在岱庄矿涌水防治上具有一定的指导意义。

Because the water quality of No.10 down limestone aquifer,No.13 limestone aquifer,Ordovician limestone equifer in Daizhuang mine has no obvious difference elements according to the current routine water quality testing data,it is easy to cause misjudgment.In order to realize the accurate identification of the inflow water source of Daizhuang mine.On the basis of collecting geological and hydrogeological data of the mine.Firstly,the hydrochemical characteristics of the main inflow aquifers are analyzed.Secondly,select Na^(+)+K^(+),Ca^(2+),Mg^(2+),Cl^(-),SO_(4)^(2-),HCO_(3)^(-)as the factor of Fisher discriminant analysis.Then use SPSS25.0 software to establish Fisher discriminant function model of inflow water source in Daizhuang mine.Finally,three known water samples from the mine were selected to verify the model.The results show that the water inflow source identification model established by Fisher′s discriminant analysis theory can accurately identify these inflow water sources with similar water quality,and the accuracy rate of back generation is 91.3%.The discrimination results of the three water samples to be judged are also consistent with the actual situation of the project,indicating that the model has certain guiding significance in the prevention and control of water inflow in Daizhuang mine.