2</sup>X(cum) 達(dá)到 99.96% ,交叉驗(yàn)證累積預(yù)測(cè)力 Q<sup>2</sup>(cum) 為 73.7% ,模型對(duì)砂巖水、太灰水和寒灰水的分類(lèi)預(yù)測(cè)準(zhǔn)確,該方法判別出的采空區(qū)水源與實(shí)際情況相符。研究為井下注漿支護(hù)和回采設(shè)計(jì)提供了理論支持,并為礦井突水的快速識(shí)別提供了一種新的定量分析方法。-龍?jiān)雌诳W(wǎng)" />

特黄三级爱爱视频|国产1区2区强奸|舌L子伦熟妇aV|日韩美腿激情一区|6月丁香综合久久|一级毛片免费试看|在线黄色电影免费|国产主播自拍一区|99精品热爱视频|亚洲黄色先锋一区

偏最小二乘回歸分析法的突水水源判別模型

  • 打印
  • 收藏
收藏成功


打開(kāi)文本圖片集

中圖分類(lèi)號(hào):TD745 文獻(xiàn)標(biāo)志碼:A 文章編號(hào):1672-9315(2025)04-0771-13

DOI:10. 13800/j. cnki. xakjdxxb. 2025.0413

Water inrush sourcediscriminationmodel based on PLS-DA algorithm

LI Ang', CHENG Xun',LV Luna1, DING Xuesong2, JING Changsheng 3 , LIU Junliang3,GAO Zhesen1, FAN Liuyi1

(1.CollegeofArchitectureandCiilEngineeing,Xi’anUniversityofScienceandTechnologyXi’anOoChina; 2.WaterResourcesand Hydropower,Hohai University,Nanjing21oo24,China; 3.Pingdingshan Tian' an Coal Industry Co., Ltd., Pingdingshan 467Ooo, China)

Abstract:To address the issues such as the insufficient integration of current discrimination methods with actual mine water source databases,a constant ion database for mine water inrush sources was developed for Pingmei No.5 Mine using a combined Piper trilinear diagram,boxplot,and cluster analysis method,Additionally,based on Partial Least Squares Determinant Analysis (PLS-DA),a comprehensive prediction and discrimination technique for mine water inrush sources was proposed. Then,a MATLABbased software system was developed to quickly and accurately identify underground water sources.34 water samples from underground aquifers were selected as training samples ,among which 6 water samples were selected as testing samples,and the model was applied in working face of Pingmei No.5 coal mine.The results indicate that: The cumulative explanatory power of the model R2X(cum) reaches 99.96%,and the cross validation cumulative predictive power Q2(cum) is 73.7% The model accurately predicts the classification of sandstone water,Taihui water,and Cambrian water,The inrush water from the gob area is consistent with field observations.This study offers theoretical support for grout reinforcement and mining progress design in underground operations,providing a new quantitative approach for rapid and precise mine water inrush source identification.

Key words: water inrush source discrimination; water source database; PLS-DA algorithm; training samples

0 引言

隨著煤礦深部工程的建設(shè)及資源開(kāi)發(fā),水害發(fā)生的頻率和嚴(yán)重程度日益增加,不僅給井下作業(yè)人員生命安全帶來(lái)巨大威脅,而且給礦山生產(chǎn)和對(duì)外供電帶來(lái)嚴(yán)重影響[1]。(剩余16085字)

monitor