• 中文核心期刊
  • 中国科技核心期刊
  • RCCSE中国核心学术期刊
  • Scopus, DOAJ, CA, AJ, JST收录期刊
高级检索

矿井涌水水源识别的Logistic回归分析方法研究

Logistic Regression Analysis Method for Identification of Mine Inflow Water Source

  • 摘要: 为了快速有效地判断识别文家坡煤矿涌水水源,利用Piper三线图和Logistic回归分析相结合的方法,在对所采集的水样水化学分析的基础上,构建了基于Logistic回归的矿井涌水水源判别模型。结果表明:通过Piper三线图能够得到待判水样的水化学特征与白垩系洛河组砂岩水和侏罗系延安组顶板砂岩水比较相似,Logistic回归模型能够排除干扰因素、准确识别待判水样1为侏罗系延安组顶板砂岩水,待判水样2为白垩系洛河组砂岩水,将两种方法相结合能够准确地识别矿井涌水水源。

     

    Abstract: In order to quickly and effectively judge and identify the source of inrush water in Wenjiapo Coal Mine, the method combining Piper trigram and Logistic regression analysis was used. On the basis of the water chemical analysis of the collected water samples, a discriminant model of mine inflow water source based on Logistic regression was established. The results show that the water chemistry characteristics of the water samples are similar to those of the Cretaceous Luohe Formation sandstone water and the Jurassic Yan'an Formation roof sandstone water. Logistic regression model can eliminate interference factors and accurately identify water sample 1 as Jurassic Yan'an Formation roof sandstone water and water sample 2 as Cretaceous Luohe Formation sandstone water. The combination of the two methods can accurately identify the source of inflow water.

     

/

返回文章
返回