基于加权的综采导水裂隙带高度多元非线性回归预测方法研究  

Height prediction of water-conducting fractured zone under fully mechanized mining based on weighted multivariate nonlinear regression

作　　者：李博[1,2,3] 吴煌 李腾 LI Bo;WU Huang;LI Teng(Key Laboratory of Karst Georesources and Environment,Ministry of Education,Guizhou University,Guiyang,Guizhou 550025,China;College of Resource and Environmental Engineering,Guizhou University,Guiyang,Guizhou 550025,China;Key Laboratory of Karst Environment and Geohazard,Ministry of La nd and Resources,Guizhou University,Guiyang,Guizhou 550025,China)

机构地区：[1]贵州大学喀斯特地质资源与环境教育部重点实验室,贵州贵阳550025 [2]贵州大学资源与环境工程学院,贵州贵阳550025 [3]贵州大学国土资源部喀斯特环境与地质灾害重点实验室,贵州贵阳550025

出　　处：《采矿与安全工程学报》2022年第3期536-545,共10页Journal of Mining & Safety Engineering

基　　金：国家自然科学基金项目(42162022,41702270);贵州省科技计划项目(黔科合基础[2019]1413,黔科合支撑[2020]4Y048);贵州省优秀青年科技人才项目(黔教合平台人才[2021]5626号)。

摘　　要：在广泛收集我国华北地区不同煤矿区的综采导水裂隙带发育高度实测数据的基础上,通过邓氏灰色关联分析理论和主成分分析理论,分析了开采厚度、开采深度、工作面斜长、硬岩岩性比例系数与导水裂隙带高度之间的关联度,并确定了各影响因素的组合权重值。采用SPSS散点分析及Matlab函数编程拟合了导水裂隙带高度与各影响因素相关关系曲线,得到了导水裂隙带发育高度对各影响因素的最佳一元非线性回归模型。在此基础上,结合各影响因素权重,构建了基于加权的综采导水裂隙带高度多元非线性回归预测修正模型,并将其与多元线性回归模型、“三下”规范经验公式的预测结果进行了对比分析,结果表明本文所建立的导水裂隙带高度预测公式准确性更高,具有较好的实际应用价值,可以为导水裂隙带高度的研究提供一定的依据和参考。Based on widely collected measured data of the height of water-conducting fractured zone under fully mechanized mining in different coal mining areas in North China, the correlation degree between mining thickness, mining depth, inclined length of panel, proportion coefficient of hard rock lithology, and the height of water-conducting fractured zone was analyzed using Deng’s grey correlation analysis and principal component analysis, and combined weight of each influencing factor was determined. SPSS scatter analysis and Matlab function programming were used to fit the correlation curve between the height of water-conducting fractured zone and influencing factors, and a most suitable one-dimensional nonlinear regression model was obtained. On this basis, the weight of each influencing factor was combined to establish a modified multivariate nonlinear regression model for height prediction of water-conducting fractured zone under fully mechanized coal mining. And the results of prediction model were compared with those of the multivariate linear regression model and the empirical formula of specification for coal mining under buildings, water and railways. Results have shown that the height prediction formula of water-conducting fractured zone established in this study has higher accuracy and better practical application, which can provide certain basis and reference for the study of the height prediction of water-conducting fractured zone.

关 键 词：导水裂隙带高度 多因素 多元非线性回归 加权 灰色关联分析 

分 类 号：TD745[矿业工程—矿井通风与安全]