基于灰熵法的深部煤层瓦斯含量影响因素分析及预测

第44卷第2期煤田地质与勘探Vol.44No.22016年4月COALGEOLOGY&EXPLORATIONApr.2016收稿日期:2015-08-12基金项目:(2015002)FoundationitemTheScientificResearchProjectofNingboUniversityofTechnology(2015002)作者简介:(1989—)[e-mail]引用格式:.[J].201644(2)19–23.ZHOUXinlongTANGJingSHIBimingetal.Analysisandforecastofinfluentialfactorsofgascontentindeepcoalseamonthebasisofthegreyentropy[J].CoalGeology&Exploration201644(2)19–23.文章编号:1001-1986(2016)02-0019-05基于灰熵法的深部煤层瓦斯含量影响因素分析及预测周鑫隆1，汤静2，石必明2,3，吕辰2(1.宁波工程学院安全工程学院，浙江宁波315016；2.安徽理工大学能源与安全学院，安徽淮南232001；3.煤矿安全高效开采省部共建教育部重点实验室，安徽淮南232001):为了提高深部煤层瓦斯含量的预测精度，提出了采用灰熵分析法对瓦斯含量影响因素进行研究，以潘三矿深部11-2煤层为例，根据灰熵关联度的大小选取不同的影响因素分别建立了GM(1,3)、GM(1,4)和GM(1,5)预测模型，依据精度检验结果选择精度更高的瓦斯预测模型。研究结果表明，影响潘三矿深部11-2煤层瓦斯含量的因素重要程度从大到小依次为：主断层距离、煤层埋深、煤厚、顶板砂泥比、煤层倾角。由此建立的3个模型的预测精度都在合格以上，其中GM(1,4)模型预测精度达到了1级，平均相对误差为5.0636%，可采用该模型对11-2煤层瓦斯含量进行预测，为深部煤与瓦斯安全高效开采提供可靠依据。：瓦斯含量；灰熵分析法；预测精度；GM(1,N)模型：P54：ADOI:10.3969/j.issn.1001-1986.2016.02.004AnalysisandforecastofinfluentialfactorsofgascontentindeepcoalseamonthebasisofthegreyentropyZHOUXinlong1，TANGJing2，SHIBiming2,3，LYUChen2(1.SchoolofSafetyEngineering,NingboUniversityofTechnology,Ningbo315016,China;2.SchoolofMiningandSafety,AnhuiUniversityofScienceandTechnology,Huainan232001,China;3.KeyLaboratoryofSafetyandHigh-efficiencyCoalMining,MinistryofEducation,AnhuiUniversityofScienceandTechnology,Huainan232001,China)Abstract:Inordertoimprovetheforecastprecisionofgascontentindeepcoalseam,takingdeepcoalseamNo.11-2inPansancoalmineasexample,greyentropyisproposedtoresearchinfluencingfactorsofgascontent.TheGM(1,3),GM(1,4)andGM(1,5)gascontentforecastingmodelsareestablishedtoselectanappropriatemodelwiththehighestforecastingprecisionaccordingtothesizeofdifferentinfluencingfactorsofgreyentropyrelationdegree.TheresultsshowthattheinfluencingfactorsofgascontentindeepcoalseamNo.11-2areindecreasingorderthemainfaultdistance,theburieddepthofcoalseam,coalseamthickness,theratioofthesandstoneandmudstoneincoalseamroofanddip.TheforecastprecisionofGMmodelsishigherthanthequalifiedlevel.What’smore,theprecisionofGM(1,4)modelreachesthefirstgradeandaveragerelativeerroris5.0636%.Inconclusion,GM(1,4)modelcanbeadoptedtoaccuratelyforecastgascontentindeepcoalseamNo.11-2,whichprovidesreliablereferencesforsafeandhigh-efficiencycoalmining.Keywords:gascontent;greyentropy;forecastprecision;GM(1,N)model[1]700~1000m·20·煤田地质与勘探第44卷[2-3][3][4-5]1灰熵关联分析法[6-7]1.1[6]x0(0)xi(0)(x1,x2,,xn)(0)(0)(0)(0)0000(0)(0)(0)(0)1111(0)(0)(0)(0)2222(0)(0)(0)(0)((1),(2),,())((1),(2),,())((1),(2),,())((1),(2),,())mmmmxxxxnxxxxnxxxxnxxxxn[8]x0(1)(k)xi(1)(k)1.2[9]minmax(1)(1)0max()()()iikxkxk(1)ΔmaxΔmin(1)(1)max0maxmax()()iikxkxk(2)(1)(1)min0minmin()()iikxkxk(3)ζ0ζ10.5Δmax(1)ζΔmaxζζΔmaxζ[10-11]ζ(1)(1)0111()()mniiikxkxkmn(4)maxiζmaxmax3,1.53,1.52ii≤≤≤≤(5)1.3Pi(k)[5]1()()()iiniikPkk(6)xi1()ln()niiiiHPkPk(7)ximax()iiiHExH(8)Hmax=lnnn2深部煤层瓦斯含量影响因素灰熵分析2.1,[12]111-2[13]2.21x0x1x2x3x4x5x0(1)(k)xi(1)(k)(1)–(5)Δmax=1.825325Δmin=0.006637ζ0.16(1)2(6)3第2期周鑫隆等:基于灰熵法的深部煤层瓦斯含量影响因素分析及预测·21·表1潘三矿11-2煤层瓦斯参数表Table1GasparametersofcoalseamNo.11-2inPansanmine/(m3·t–1)(x1)/m(x2)/m(x3)(x4)/m(x5)/(°)16.32673.1840.30.9721.881027.29726.76673.390.7631.91037.21705.28891.70.9071.86746.9680.23807.471.011.78556.96685.81867.850.8691.81766.74704.18739.710.8941.9876.78661.65757.330.8081.85586.29672.41800.421.0951.73896.24639.13751.580.9821.895106.19655.75780.720.9021.665116.46697.13773.831.0351.7515127.21705.28891.70.9071.867表2灰关联系数表Table2Greyrelationcoefficient123456789101112δ10.5710.7260.8330.7820.8750.9571.0000.8920.8710.6830.6910.619δ20.6410.7250.9030.8260.9910.8910.7570.8100.9480.9160.9290.953δ30.5030.3410.3550.5210.7120.6590.8930.7650.5280.3610.3650.329δ40.5510.6220.7370.7700.9390.8920.9110.8450.8260.6380.6130.634δ50.9930.5810.8920.5170.1410.5170.4460.6110.4680.3350.3340.450表3灰色关联系数分布映射密度值Table3Densitymappedbythedistributionofgreyrelationcoefficient123456789101112P10.0600.0760.0880.0820.0920.1010.1050.0940.0920.0720.0730.065P20.0620.0710.0880.0800.0960.0870.0740.0790.0920.0890.0900.093P30.0790.0540.0560.0820.1120.1040.1410.1210.0830.0570.0580.052P40.0610.0690.0820.0860.1050.0990.1010.0940.0920.0710.0680.071P50.1580.0920.1420.0820.0220.0820.0710.0970.0740.0530.0530.072(7)4(8)1表4各影响因素灰关联熵Table4Greyrelationentropyofinfluentialfactorsx1x2x3x4x52.47142.47762.42692.47012.39371.5351.5401.5081.5321.48513深部煤层瓦斯含量灰色预测3.1GM(1,N)(AGO)“”GM[14-15]GM(1,N)1N[16]xi(0)(k)nNi=1,2,,Nk=1,2,,n1-AGOxi(1)(k)xi(1)(k)[17]·22·煤田地质与勘探第44卷(1)(1)(1)112ddNiiixaxbxt(9)ˆa=[a,b1,b2,,bN]TT1Tˆ()NaBBBY,(1)(1)(1)(1)112(1)(1)(1)(1)112(1)(1)(1)(1)1121((1)(2))(2)(2)21((2)(3))(3)(2)21((1)())()(2)2NNNxxxxxxxxxnxnxnxB(10)(0)(0)(0)111(2)(3)()TNYxxxn(11)GM(1,N)[7](1)(1)(1)112(1)21(1)((0)(1))e1(1)NakiiiNiiixkxbxkabxka(12)x1(1)(0)=x1(0)(1)[7]0(1)(1)11111xkxkxk(13)[18]5cp6表5误差检验计算公式Table5Formulasforcheckingcomputationofresidualerrors(0)11niin(0)11niiXXn2(0)111()niiSn2(0)211()niiSXXnc12ScSp(