基于机器学习的煤接触角预测模型研究

煤的接触角不仅可以表征煤的润湿性，还可以表征煤的可浮性，当煤接触角越小时，煤的润湿性越好，煤的可浮性越差。就工业分析、元素分析、孔隙结构等对接触角的影响，国内外学者已做了大量的研究分析，但煤的矿物组成对接触角的影响研究报道尚不多见，同时在煤接触角的量化模型研究中往往由于数据较少而难以对实际煤样的接触角做出较为可靠的估测。因此，研究煤的矿物组成对接触角的影响，构建可用于实际煤样接触角估计的量化模型具有重要学术意义。
本文在总结分析前人有关煤接触角的影响因素及量化模型研究工作的基础上，提出了一种基于机器学习的煤接触角预测方法：（1）按有机质、高岭石、石英和黄铁矿之间的不同质量比调配煤样，通过胶体磨研磨混合，并进行接触角测定获得人工调制煤样组成及接触角数据；（2）建立支持向量机和随机森林模型，并在剔除异常数据后采用接触角均值数据和原始数据训练模型；（3）网格搜索确定模型参数，使用模型预测样品的接触角，并给出预测的可靠程度。
本文对机器学习及其在煤炭加工行业的应用研究进行了简要介绍，并对随机森林和支持向量机的原理进行了阐述。为满足接触角预测模型的训练和测试要求，将实验数据划分成训练集和测试集。针对模型的性能评判，引入4 个评价指标检验模型的预测效果。为获得模型的最佳参数组合，使用网格搜索法对参数进行优化。
本文构建了基于随机森林的煤接触角预测模型和基于支持向量机的煤接触角预测模型，并对比了模型性能，结果表明，在剔除异常数据后的原始样本上基于随机森林的煤接触角预测模型性能最好，在测试集和训练集上表现相当，在测试集上判定系数达到0.9225，具有较高准确率。将其用于真实煤样的接触角预测，预测结果较准确，表明该模型可用于评估真实煤样接触角。

[1] 谢和平, 任世华, 谢亚辰, 等. 碳中和目标下煤炭行业发展机遇[J]. 煤炭学报, 2021, 46(7): 2197-2211.
[2] 中华人民共和国国家统计局. 中国统计年鉴.2021[M]. 北京: 中国统计出版, 2021: 178-178.
[3] 蔡璋, 蒋荣立, 罗时磊, 等. 煤泥的选择性絮凝研究[J]. 煤炭学报, 1994, 19(5): 513-520.
[4] 谢广元, 吴玲, 欧泽深, 等. 从细粒煤泥中回收精煤的分选与脱水技术研究[J]. 煤炭学报, 2004, 29(5): 602-605.
[5] 董宪姝, 杜圣星. 高灰细泥细粒煤浮选技术进展[J]. 选煤技术, 2012(5): 110-114.
[6] WILLS B A, FINCH J A. Wills’ mineral processing technology: an introduction to the practical aspects of ore treatment and mineral recovery[M]. Butterworth-Heinemann, 2015: 267-270.
[7] SOBHY A, TAO D. Nanobubble column flotation of fine coal particles and associated fundamentals[J]. International Journal of Mineral Processing, 2013, 124: 109-116.
[8] FENG D, ALDRICH C. Effect of particle size on flotation performance of complex sulphideores[J]. Minerals Engineering, 1999, 12(7): 721-731.
[9] 魏凌敖. 基于机器视觉的煤泥浮选加药控制系统研究[D]. 中国矿业大学, 2020.
[10] 刘敏, 张友飞, 郭芳余, 等. 表面粗糙度对煤泥可浮性的影响[J]. 煤炭科学技术, 2019, 47(10): 253-258.
[11] 史涛涛. 有关煤粉(泥) 可浮性评定方法的一些看法[J]. 煤炭加工与综合利用, 2014(3): 18-21.
[12] 贺萌, 由晓芳, 张伟, 等. 非离子表面活性剂在低阶煤表面的吸附特性及其对润湿性的影响[J]. 中国科技论文, 2017, 12(15): 1704-1710.
[13] XU G, CHEN Y P, EKSTEEN J, et al. Surfactant-aided coal dust suppression: a review of evaluation methods and influencing factors[J]. Science of The Total Environment, 2018, 639:1060-1076.
[14] DEGANELLO D, CROFT T, WILLIAMS A, et al. Numerical simulation of dynamic contact angle using a force based formulation[J]. Journal of Non-Newtonian Fluid Mechanics, 2011, 166(16): 900-907.
[15] YOUNG T. III. An essay on the cohesion of fluids[J]. Philosophical Transactions of The Royal Society of London, 1805(95): 65-87.
[16] 谢广元. 选矿学[M]. 徐州: 中国矿业大学出版社, 2001: 400-402.
[17] 郭志强. 桥联改性活化浮选硅孔雀石研究[D]. 昆明理工大学, 2019.
[18] MILLER J D, NALASKOWSKI J, ABDUL B, et al. Surface characteristics of kaolinite and other selected two layer silicate minerals[J]. The Canadian Journal of Chemical Engineering, 2007, 85(5): 617-624.
[19] 王延秋. 表面活性剂对煤尘润湿效果的定量化研究[D]. 西安科技大学, 2020.
[20] 谢克昌. 煤的结构与反应性[M]. 北京: 科学出版社, 2002: 1-67.
[21] 李娇阳. 煤表面润湿性影响因素分析[D]. 河南理工大学, 2016.
[22] 彭扬东, 石彦平, 陈书雅, 等. 煤岩矿物组成与微观结构对其润湿性的影响规律研究[J]. 煤炭技术, 2018, 37(7): 112-114.
[23] 程卫民, 薛娇, 周刚, 等. 烟煤煤尘润湿性与无机矿物含量的关系研究[J]. 中国矿业大学学报, 2016, 45(3): 462-468.
[24] 张建国, 李红梅, 刘依婷, 等. 煤尘微细观润湿特性及抑尘剂研发初探——以平顶山矿区为例[J]. 煤炭学报, 2021, 46(3): 812-825.
[25] 孙银宇. 煤尘润湿性研究及降尘剂复配方案[D]. 安徽理工大学, 2014.
[26] CHEN Y, MA D M, XIA Y C, et al. Study on wettability and influencing factors of different macroscopic components in low rank coal[J]. Coal Science and Technology, 2019, 47(9): 97-104.
[27] 周刚, 薛娇, 程卫民, 等. 基于X 射线衍射实验的堆垛结构对煤尘润湿性的影响[J]. 工程科学学报, 2015, 37(12): 1535-1541.
[28] WANG X N, YUAN S J, JIANG B Y. Experimental investigation of the wetting ability of surfactants to coals dust based on physical chemistry characteristics of the different coal samples[J]. Advanced Powder Technology, 2019, 30(8): 1696-1708.
[29] GOSIEWSKA A, DRELICH J, LASKOWSKI J, et al. Mineral matter distribution on coal surface and its effect on coal wettability[J]. Journal of Colloid and Interface Science, 2002, 247(1): 107-116.
[30] MAHONEY S A, RUFFORD T E, DMYTERKO A S, et al. The effect of rank and lithotype on coal wettability and its application to coal relative permeability models[C]//SPE Asia Pacific Unconventional Resources Conference and Exhibition. OnePetro, 2015: 1-10.
[31] CRAWFORD R J, GUY D W, MAINWARING D E. The influence of coal rank and mineral matter content on contact angle hysteresis[J]. Fuel, 1994, 73(5): 742-746.
[32] GUTIERREZ-RODRIGUEZ J, PURCELL JR R, APLAN F. Estimating the hydrophobicity of coal[J]. Colloids and Surfaces, 1984, 12(1): 1-25.
[33] 杨静, 徐辉, 高建广, 等. 粒度对煤尘表面特性及润湿性的影响[J]. 煤矿安全, 2014, 45(10): 140-143.
[34] LI Q Z, LIN B Q, ZHAO S, et al. Surface physical properties and its effects on the wetting behaviors of respirable coal mine dust[J]. Powder Technology, 2013, 233: 137-145.
[35] 谭烜昊, 王鹏飞, 易波波, 等. 煤尘润湿性能与粒径关系的实验研究[J]. 矿业工程研究, 2018, 33(2): 14-17.
[36] WANG P F, TAN X H, ZHANG L Y, et al. Influence of particle diameter on the wettability of coal dust and the dust suppression efficiency via spraying[J]. Process Safety and Environmental Protection, 2019, 132: 189-199.
[37] 张新花, 徐翠翠, 颜国强, 等. 不同煤种润湿性影响因素分析[J]. 煤矿安全, 2015, 46(1): 156-158.
[38] 安文博, 王来贵. 表面活性剂作用下煤体力学特性及改性规律[J]. 煤炭学报, 2020, 45(12):4074-4086.
[39] MA Y L, ZHU X W. Mechanism of surfactant improve water wetting coal dust[J]. Coal Technology, 2015, 34(5): 195-198.
[40] XI X, JIANG S G, ZHANG W Q, et al. An experimental study on the effect of ionic liquids on the structure and wetting characteristics of coal[J]. Fuel, 2019, 244: 176-183.
[41] TIEN J C, KIM J. Respirable coal dust control using surfactants[J]. Applied Occupational and Environmental Hygiene, 1997, 12(12): 957-963.
[42] ORUMWENSE F O. Estimation of the wettability of coal from contact angles using coagulants and flocculants[J]. Fuel, 1998, 77(9): 1107-1111.
[43] 宋继伟, 彭扬东, 石彦平, 等. 贵州六盘水煤样润湿性影响因素研究[J]. 矿业安全与环保, 2018, 45(1): 16-19.
[44] XIA W C, NIU C K, LI Y F. Effect of heating process on the wettability of fine coals of various ranks[J]. The Canadian Journal of Chemical Engineering, 2017, 95(3): 475-478.
[45] SAVITSKYI D. Impact of the pH on angles of contact of water wettability of brown coal[J]. Journal of Water Chemistry and Technology, 2015, 37(4): 155-160.
[46] CHATURVEDI T, SCHEMBRE J, KOVSCEK A. Spontaneous imbibition and wettability characteristics of Powder River Basin coal[J]. International Journal of Coal Geology, 2009, 77(1): 34-42.
[47] DING C, NIE B S, YANG H, et al. Experimental research on optimization and coal dust suppression performance of magnetized surfactant solution[J]. Procedia Engineering, 2011, 26: 1314-1321.
[48] 马艳玲. 新型煤尘润湿剂的实验研究[D]. 安徽理工大学, 2016.
[49] 文金浩, 薛娇, 张磊, 等. 基于XRD 分析长焰煤润湿性与其灰分的关系[J]. 煤炭科学技术, 2015, 43(11): 83-86.
[50] 罗根华, 李博, 丁莹莹, 等. 煤尘化学组成及结构参数对煤尘润湿性的影响规律[J]. 大连交通大学学报, 2016, 37(3): 64-67.
[51] GAO J G, YANG J. Study of coal dust wettability based on multiple stepwise regression analysis[J]. Safety in Coal Mines, 2012, 43(1): 126-129.
[52] MURATA T. Studies on wettability of coal (2nd report) the relation between wettability of coal and coal elementary composition[J]. Journal of The Mining and Metallurgical Institute of Japan, 1981, 97(1123): 937-943.
[53] CARBONELL J G, MICHALSKI R S, MITCHELL T M. An overview of machine learning[J]. Machine Learning, 1983, 1: 3-23.
[54] YELIN I, SNITSER O, NOVICH G, et al. Personal clinical history predicts antibiotic resistance of urinary tract infections[J]. Nature Medicine, 2019, 25(7): 1143-1152.
[55] ROUGH K, DAI A M, ZHANG K, et al. Predicting inpatient medication orders from electronic health record data[J]. Clinical Pharmacology and Therapeutics, 2020, 108(1): 145-154.
[56] KHANAL S, FULTON J, KLOPFENSTEIN A, et al. Integration of high resolution remotely sensed data and machine learning techniques for spatial prediction of soil properties and corn yield[J]. Computers and Electronics in Agriculture, 2018, 153: 213-225.
[57] FILIPPI P, JONES E J, WIMALATHUNGE N S, et al. An approach to forecast grain crop yield using multi-layered, multi-farm data sets and machine learning[J]. Precision Agriculture, 2019, 20(5): 1015-1029.
[58] GOUDRA B, SINGH P M. Failure of sedasys: destiny or poor design[J]. Anesthesia and Analgesia, 2017, 124(2): 686-688.
[59] MIRSADEGHI M, BEHNAM H, SHALBAF R, et al. Characterizing awake and anesthetized states using a dimensionality reduction method[J]. Journal of Medical Systems, 2016, 40(1): 1-8.
[60] ASCARZA E, NETZER O, HARDIE B G. Some customers would rather leave without saying goodbye[J]. Marketing Science, 2018, 37(1): 54-77.
[61] LU S, XIAO L, DING M. A video-based automated recommender (VAR) system for garments[J]. Marketing Science, 2016, 35(3): 484-510.
[62] 任浩. 基于BP 神经网络的浓缩机药剂添加系统设计与应用[J]. 能源与节能, 2018(12): 178-179.
[63] 王靖千, 王然风, 付翔, 等. 基于彩色图像处理的浮选尾煤灰分软测量研究[J]. 煤炭工程, 2020, 52(3): 137-142.
[64] 王光辉. 煤泥浮选过程模型仿真及控制研究[D]. 中国矿业大学, 2012.
[65] 曹文龙. 基于LabVIEW 的煤泥浮选泡沫图像处理系统研究[D]. 中国矿业大学, 2016.
[66] 杨晓鸿, 郑诚, 王昊鑫. 浮选加药量预测模型的研究[J]. 选煤技术, 2020(1): 87-90.
[67] 曹珍贯. 重介选煤过程中重介质的密度预测控制研究[D]. 中国矿业大学, 2014.
[68] JORJANI E, MESROGHLI S, CHELGANI S C. Prediction of operational parameters effect on coal flotation using artificial neural network[J]. Journal of University of Science and Technology Beijing, 2008, 15(5): 528-533.
[69] ALI D, HAYAT M B, ALAGHA L, et al. An evaluation of machine learning and artificial intelligence models for predicting the flotation behavior of fine high-ash coal[J]. Advanced Powder Technology, 2018, 29(12): 3493-3506.
[70] RAO B V, GOPALAKRISHNA S. Hardgrove grindability index prediction using support vector regression[J]. International Journal of Mineral Processing, 2009, 91(1): 55-59.
[71] LI P S, XIONG Y H, YU D X, et al. Prediction of grindability with multivariable regression and neural network in Chinese coal[J]. Fuel, 2005, 84(18): 2384-2388.
[72] CEYLAN Z, SUNGUR B. Estimation of coal elemental composition from proximate analysis using machine learning techniques[J]. Energy Sources Part A-Recovery Utilization and Environmental Effects, 2020, 42(20): 2576-2592.
[73] BREIMAN L. Random forests[J]. Machine Learning, 2001, 45(1): 5-32.
[74] CUTLER A, CUTLER D R, STEVENS J R. Random forests[J]. Ensemble Machine Learning, 2012: 157-175.
[75] CORTES C, VAPNIK V. Support-vector networks[J]. Machine Learning, 1995, 2(3): 273-297.
[76] BHATI B S, RAI C. Analysis of support vector machine-based intrusion detection techniques[J]. Arabian Journal for Science and Engineering, 2020, 45(4): 2371-2383.
[77] ZHU H, LIU X, LU R, et al. Efficient and privacy-preserving online medical prediagnosis framework using nonlinear SVM[J]. IEEE Journal of Biomedical and Health Informatics, 2016, 21(3): 838-850.
[78] 宋党育. 煤中矿物质的定量及赋存特征研究[M]. 徐州: 中国矿业大学出版社, 2011: 5-6.
[79] 雷绍民. 高岭石基纳米TiO2 复合光催化材料研究[D]. 武汉理工大学, 2006.
[80] 戴新枝. 浅谈350MW 机组混煤燃烧技术[J]. 上海节能, 2021(3): 303-308.
[81] 陈文敏. 煤质分析结果的定性与定量审查[M]. 北京: 煤炭工业出版社, 1994: 7-23.
[82] 史兴民. 控制煤中SO2 排放的技术对策分析[J]. 资源开发与市场, 2006, 22(5): 425-428.
[83] 刘蒙蒙. 乌鲁木齐矿区大倾角地层煤层气钻井井壁稳定性研究[J]. 中国煤层气, 2020, 17(4): 24-28.
[84] 郭明明. 煤泥浮选过程中气泡与煤及矿物质粘附几率的研究[D]. 太原理工大学, 2017.
[85] BREIMAN L. Bagging predictors[J]. Machine Learning, 1996, 24(2): 123-140.
[86] TIBSHIRANI R. Bias, variance and prediction error for classification rules[M]. Citeseer, 1996: 1-17.
[87] QI X, LI X, LIANG Y, et al. Surface structure-dependent hydrophobicity/oleophilicity of pyrite and its influence on coal flotation[J]. Journal of Industrial and Engineering Chemistry, 2020, 87: 136-144.
[88] 周宝楠. 微细颗粒分离过程强化研究[D]. 哈尔滨工业大学, 2019.