摘要
为解决煤矿瓦斯涌出量预测过程中存在的指标繁杂致使预测精度低的问题,首先运用R语言中的主成分分析法(PCA)对煤矿瓦斯涌出量的影响因素进行降维分析;然后引入量纲分析理论,采用无量纲化处理方法消除量纲对数据的影响;最后利用多元回归分析法(MRA)结合多种多元回归模型,最终选用多元线性回归预测模型对煤矿瓦斯涌出量进行预测,并选取部分数据对所建立的煤矿瓦斯涌出量预测模型进行验证。结果表明:主成分分析法能有效减少预测变量的个数,经主成分旋转后构建的多元回归预测模型的预测精度较高,其相对误差平均值为1.99%;主成分分析与多元线性回归分析相结合的方法适用于煤矿瓦斯涌出量的预测,且满足工程需要,可为井下瓦斯涌出量预测技术的选择提供参考。
In order to deal with the poor prediction accurac of the coal mine gas emission for the existing comlicated prediction index,this article analyses the influencing factor of gas emission by using principal component analysis method in R language.Then the paper introduces dimension analysis theory,and uses dimensionless processing to eliminate the influence of dimension on data.At the end the paper applies the multivariate linear regression combined with a variety of multiple regression models to predicting the gas emission in coal min.The paper selects some data to verify the predicted function of gas emission.The results show that the principal component analysis method can effectively reduce the number of predictors,the multivariate regression prediction model constructed by the principal component rotation has higher precision,and the average value of its relative error is 1.99%.The combination of principal component analysis and multiple linear regression is applicable to the prediction of gas emission in coal mine,meets the engineering needs.This method can provide some reference for the selection of underground gas emission prediction technology.
引文
[1] 何清.工作面瓦斯涌出量预测研究现状及发展趋势[J].矿业安全与环保,2016,43(4):98-101.
[2] Creedya D P.Geological controls on the formation and distribution of gas in British coal measure strata[J].International Journal of Coal Geology,1988,10(1):1-31.
[3] 瓦斯通风防灭火安全研究所.矿井瓦斯涌出量预测方法的发展与贡献[J].煤矿安全,2003,34(增刊):10-13.
[4] 戴永禄.2种瓦斯涌出预测法在屯兰矿的应用对比[J].煤炭科学技术,2011(专刊):30-32.
[5] 李晓华,周炳秋,韩真理.瓦斯地质及瓦斯涌出动态预测可视化系统研究及应用[J].煤炭工程,2016,48(4):62-64.
[6] 张占国,侯志华.基于分源预测法的厚煤层分层开采瓦斯治理技术研究[J].矿业安全与环保,2017,44(1):78-82.
[7] 张子戌,袁崇孚.瓦斯地质数学模型法预测矿井瓦斯涌出量研究[J].煤炭学报,1999,24(4):368-372.
[8] Dong D W.Mine Gas emission prediction based on gaussian process model[J].Procedia Engineering,2012,45(2):334-338.
[9] 董晓雷,贾进章,白洋,等.基于SVM耦合遗传算法的回采工作面瓦斯涌出量预测[J].安全与环境学报,2016,16(2):114-118.
[10]Li R Q,Shi S L,Wu A Y,et al.Research on prediction of gas emission based on self-organizing data mining in coal mines[J].Procedia Engineering,2014,84:779-785.
[11]景国勋,衡献伟.基于灰色预测模型与一元线性回归模型的煤矿瓦斯涌出量预测比较[J].安全与环境工程,2010,17(5):63-65.
[12]Liang C,Wang E Y,Feng J J,et al.A dynamic gas emission prediction model at the heading face and its engineering application[J].Journal of Natural Gas Science and Engineering,2016,30:228-236.
[13]Booth P,Brown H,Nemcik J,et al.Spatial context in the calculation of gas emissions for underground coal mines[J].International Journal of Mining Science and Technology,2017,27(5):787-794.
[14]巴科弗.R语言实战[M].高涛,肖楠,陈钢,译.2版.北京:人民邮电出版社,2013.
[15]朱红青,常文杰,张彬.回采工作面瓦斯涌出BP神经网络分源预测模型及应用[J].煤炭学报,2007,32(5):504-508.
[16]赵国彦,梁伟章,王少锋,等.基于量纲分析的巷道围岩松动圈预测模型[J].岩土力学,2016,37(增刊2):273-278,300.