Random forest method for predicting coal spontaneous combustion in gob
-
摘要: 煤自然发火温度的准确预测是矿井煤自燃防控的关键。为了准确可靠地预测采空区煤自燃温度,在大佛寺煤矿40106综放工作面开展了长期的采空区温度和气体观测实验,提出了一种基于随机森林(RF)方法的采空区煤自燃预测模型,并将预测结果与支持向量机(SVM)和BP神经网络(BPNN)方法对比。采用粒子群优化算法(PSO)对RF和SVM超参数进行优化,建立了参数优化的PSO-RF和PSO-SVM预测模型。结果表明,RF,PSO-RF,SVM和PSO-SVM模型均具有较强的泛化性和鲁棒性;RF在建模过程中拥有宽广的参数适应范围,当树的数量(ntree)超过100后,其训练误差趋于稳定,ntree的改变对预测性能没有实质的影响;虽然PSO算法可以找到RF最优超参数,但默认参数的RF模型就能获得满意的预测性能;SVM预测结果则对超参数十分敏感,PSO优化可以显著提高其预测精度,其预测性能依赖于超参数的最优选择;BPNN模型在训练阶段拥有极佳的预测结果,但易出现“过拟合”,导致泛化性弱,测试阶段误差较大。通过在其他矿井煤自燃预测中应用,验证了RF方法的稳定性和普适性,且无需复杂参数设置和优化就能获得良好的预测性能,可进一步应用于其他能源燃料领域。Abstract: The accurate prediction of coal temperature plays a vital role in preventing and controlling the coal spontaneous combustion in coal mines.To predict the temperature of coal spontaneous combustion in a gob accurately and reliably, a long-term observation test of temperature and gases was implemented in the gob of 40106 fully mechanized top-coal caving face at Dafosi coal mine.A prediction model of coal spontaneous combustion in the gob based on random forest ( RF) method was proposed, and the prediction results were compared with the support vector machine ( SVM) and BP neural network ( BPNN) methods.The particle swarm optimization ( PSO) algorithm was employed to optimize the hyper-parameters of RF and SVM for establishing the PSO-RF and PSO-SVM prediction models with optimized parameters.The results indicate that RF, PSO-RF, SVM, and PSO-SVM models all had strong generalization androbustness.RF possessed a wide range of parameters adaptation in the modeling process. When the number of trees (ntree) exceeded 100, the training errors tended to be stable, and the change of ntreehad no substantial impact on the prediction performance.Although the PSO algorithm could find the optimal hyper-parameters of RF, the RF model with the default parameters could obtain a satisfactory prediction performance.The prediction results of SVM were very sensitive to its hyper-parameters, PSO optimization could significantly improve its prediction accuracy, and its prediction performance depended on the optimal choice of hyper-parameters.The BPNN model exhibited excellent prediction results in the training stage, but it was prone to “over-fitting”, resulting in weak generalization and large errors in the testing stage.Through the application of coal spontaneous combustion prediction in other mines, the stability and universality of the RF method were verified, and good prediction performance could be obtained without complicated parameter settings and optimization, it could be further applied to other energy and fuel fields.
-
-
[1] SONG Z Y, KUENZER C.Coal fires in China over the last decade:A comprehensive review[J]. International Journal of Coal Geology, 2014, 133:72-99.
[2] SAFFARI A, SERESHKI F, ATAEI M, et al.Presenting an engineering classification system for coal spontaneous combustion potential[J]. International Journal of Coal Science&Technology, 2017, 4 (2): 110-128.
[3] DENG J, LEI C K, XIAO Y, et al.Determination and prediction on“three zones”of coal spontaneous combustion in a gob of fully mechanized caving face[J].Fuel, 2018, 211:458-470.
[4] 李林, 陈军朝, 姜德义, 等.煤自燃全过程高温区域及指标气体时空变化实验研究[J].煤炭学报, 2016, 41 (2): 444-450. LI Lin, CHEN Junchao, JIANG Deyi, et al. Experimental study on temporal variation of high temperature region and index gas of coal spontaneous combustion[J]. Journal of China Coal Society, 2016, 41 (2): 444-450.
[5] HU X C, YANG S Q, ZHOU X H, et al.Coal spontaneous combustion prediction in gob using chaos analysis on gas indicators from upper tunnel[J]. Journal of Natural Gas Science and Engineering, 2015, 26:461-469.
[6] 邓军, 李贝, 李珍宝, 等.预报煤自燃的气体指标优选试验研究[J].煤炭科学技术, 2014, 42 (1): 55-59. DENG Jun, LI Bei, LI Zhenbao, et al.Experiment study on gas indexes optimization for coal spontaneous combustion prediction[J].Coal Science and Technology, 2014, 42 (1): 55-59.
[7] DENG J, MA X F, ZHANG Y T, et al.Effects of pyrite on the spontaneous combustion of coal[J].International Journal of Coal Science&Technology, 2015, 2 (4): 306-311.
[8] 王磊, 武术静, 李长青.基于修正灰色马尔科夫模型的煤自燃预测[J].计算机仿真, 2014, 31 (11): 416-420. WANG Lei, WU Shujing, LI Changqing. Coal spontaneous combustion prediction based on grey-Markov model[J]. Computer Simulation, 2014, 31 (11): 416-420.
[9] 邬剑明, 王俊峰.基于神经网络的煤层自然发火的非线性预测[J].中国安全科学学报, 2004, 14 (5): 15-17. WU Jianming, WANG Junfeng.A non-linear prediction of coal selfignition based on neural network[J]. China Safety Science Journal, 2004, 14 (5): 15-17.
[10] 周福宝, 李金海.煤矿火区启封后复燃预测的BP神经网络模型[J].采矿与安全工程学报, 2010, 27 (4): 494-498, 504. ZHOU Fubao, LI Jinhai.Prediction model for reignition of fire zone after unsealing based on BP neural networks[J].Journal of Mining and Safety Engineering, 2010, 27 (4): 494-498, 504.
[11] 靳玉萍.基于代数神经网络的煤自燃预测[J].数学的实践与认识, 2013, 43 (18): 122-128. JIN Yuping. Forecasting of coal spontaneous combustion based on algebra neural networks[J]. Mathematics in Practice and Theory, 2013, 43 (18): 122-128.
[12] 高原, 覃木广, 李明建.基于支持向量机的采空区遗煤自燃预测分析[J].煤炭科学技术, 2010, 38 (2): 50-54. GAO Yuan, QIN Muguang, LI Mingjian. Analysis on prediction of residual coal spontaneous combustion in goaf based on support vector machine[J]. Coal Science and Technology, 2010, 38 (2): 50-54.
[13] 孟倩, 王永胜, 周延.基于粗集-支持向量机的采空区自然发火预测[J].煤炭学报, 2010, 35 (12): 2100-2104. MENG Qian, WANG Yongsheng, ZHOU Yan. Prediction of spontaneous combustion in caving zone based on rough set and support vector machine[J].Journal of China Coal Society, 2010, 35 (12): 2100-2104.
[14] 邵良杉, 李相辰.不平衡数据下的采空区自然发火预测研究[J].中国安全科学学报, 2017, 27 (6): 61-66. SHAO Liangshan, LI Xiangchen. Prediction of coal spontaneous combustion in caving zone with unbalanced data[J]. China Safety Science Journal, 2017, 27 (6): 61-66.
[15] 靳玉萍, 张兵, 高凯.球形支持向量机在煤自燃预测中的应用[J].计算机应用与软件, 2013, 30 (9): 57-60. JIN Yuping, ZHANG Bing, GAO Kai.Application of spheical-SVM in forecasting spontaneous combustion of coal[J].Computer Applications and Software, 2013, 30 (9): 57-60.
[16] 邓军, 周少柳, 马砺, 等.基于PCA-PSOSVM的煤自燃程度预测研究[J].矿业安全与环保, 2016, 43 (5): 27-31. DENG Jun, ZHOU Shaoliu, MA Li, et al. Research on prediction method of coal spontaneous combustion degree based on PCAPSOSVM[J]. Mining Safety and Environmental Protection, 2016, 43 (5): 27-31.
[17] 邓军, 雷昌奎, 曹凯, 等.煤自燃预测的支持向量回归方法[J].西安科技大学学报, 2018, 38 (2): 175-180. DENG Jun, LEI Changkui, CAO Kai, et al.Support vector regression approach for predicting coal spontaneous combustion[J].Journal of Xi’an University of Science and Technology, 2018, 38 (2): 175-180.
[18] BREIMAN L.Random forests[J].Machine Learning, 2001, 45 (1): 5-32.
[19] MATIN S S, CHELGANI S C.Estimation of coal gross calorific value based on various analyses by random forest method[J]. Fuel, 2016, 177:274-278.
[20] CHELGANI S C, MATIN S S, MAKAREMI S. Modeling of free swelling index based on variable importance measurements of parent coal properties by random forest method[J]. Measurement, 2016, 94:416-422.
[21] CHELGANI S C, MATIN S S, HOWER J C. Explaining relationships between coke quality index and coal properties by random forest method[J].Fuel, 2016, 182:754-760.
[22] MATIN S S, HOWER J C, FARAHZADI L, et al.Explaining relationships among various coal analyses with coal grindability index by random forest[J]. International Journal of Mineral Processing, 2016, 155:140-146.
[23] BREIMAN L, FRIEDMAN J H, OLSHEN R A, et al.Classification and regression trees[M]. Belmont (California): Wadsworth International Group, 1984.
[24] CORTES C, VAPNIK V. Support-vector networks[J]. Machine Learning, 1995, 20:273-297.
[25] 陈同俊, 王新, 管永伟.基于SVR和地震属性的构造煤厚度定量预测[J].煤炭学报, 2015, 40 (5): 1103-1108. CHEN Tongjun, WANG Xin, GUAN Yongwei. Quantitative prediction of tectonic coal seam thickness using support vector regression and seismic attributes[J]. Journal of China Coal Society, 2015, 40 (5): 1103-1108.
[26] CHANG C C, LIN C J.LIBSVM:A library for support vector machines[J].ACM Transactions on Intelligent Systems and Technology, 2011, 2 (3): 1-27.
[27] 施龙青, 谭希鹏, 王娟, 等.基于PCA_Fuzzy_PSO_SVC的底板突水危险性评价[J].煤炭学报, 2015, 40 (1): 167-171. SHI Longqing, TAN Xipeng, WANG Juan, et al. Risk assessment of water inrush based on PCA_Fuzzy_PSO_SVC[J].Journal of China Coal Society, 2015, 40 (1): 167-171.
[28] 李慧民, 李振雷, 何荣军, 等.基于粒子群算法和BP神经网络的冲击危险性评估[J].采矿与安全工程学报, 2014, 31 (2): 203-207, 231. LI Huimin, LI Zhenlei, HE Rongjun, et al. Rock burst risk evaluation based on particle swarm optimization and BP neural network[J].Journal of Mining and Safety Engineering, 2014, 31 (2): 203-207, 231.
[29] LIAW A, WIENER M.Classification and regression by random forest[J].R News, 2002, 2:18-22.
[30] XIE Z H, CAI J, ZHANG Y. Division of spontaneous combustion“three-zone”in goaf of fully mechanized coal face with big dip and hard roof[A].2012 international symposium on safety science and engineering in China[C].Procedia Engineering, 2012:82-87.
计量
- 文章访问数: 836
- HTML全文浏览量: 1
- PDF下载量: 187