基于随机森林回归算法的回采工作面瓦斯涌出量预测

张增辉; 马文伟

doi:10.13272/j.issn.1671-251x.2023020006

基于随机森林回归算法的回采工作面瓦斯涌出量预测

doi: 10.13272/j.issn.1671-251x.2023020006

张增辉^1,,
马文伟^{2, 3, 4}

1.
国家能源集团神东煤炭集团有限责任公司保德煤矿，山西忻州　036600
2.
中煤科工集团沈阳研究院有限公司，辽宁抚顺　113122
3.
煤矿安全技术国家重点实验室，辽宁抚顺　113122
4.
西安交通大学人居环境与建筑工程学院，陕西西安　710049

基金项目: 国家科技重大专项资助项目（2016ZX05045-004-001）。

详细信息

作者简介:
张增辉（1987—），男，河南西华人，工程师，硕士，研究方向为井工煤矿灾害治理，E-mail：544396324@qq.com

中图分类号: TD712
计量
- 文章访问数: 126
- HTML全文浏览量: 63
- PDF下载量: 13
- 被引次数: 0
出版历程
- 收稿日期: 2023-07-22
- 修回日期: 2023-12-10
- 网络出版日期: 2024-01-03

Prediction of gas emission in mining face based on random forest regression algorithm

ZHANG Zenghui^1
,,
MA Wenwei^{2, 3, 4}

1.
Baode Coal Mine, CHN Energy Shendong Coal Group Co., Ltd., Xinzhou 036600, China
2.
CCTEG Shengyang Research Insititute, Fushun 113122, China
3.
State Key Laboratory of Coal Mine Safety Technology, Fushun 113122, China
4.
School of Human Settlements and Civil Engineering, Xi'an Jiaotong University, Xi'an 710049, China

摘要

摘要:
回采工作面是矿井瓦斯涌出的主要场所，精准预测回采工作面的瓦斯涌出量，进而有针对性地提出防治措施，对保证矿井安全生产具有重要意义。提出了基于随机森林回归算法的回采工作面瓦斯涌出量预测方法。以工作面实测瓦斯涌出量数据为原始样本，利用Bootstrap抽样方法进行随机抽样，以袋外数据（OOB）评估分数oob_score作为随机森林回归模型调参、特征变量重要性的评判指标，计算得出模型的最佳参数、特征变量重要性占比。对各特征变量的重要性占比进行排序，并按排序进行随机森林回归模型性能分析，结果表明：随着特征变量数的增加，模型性能不会呈现规律性的变化；当特征变量数较少时，可能存在过拟合的情况。测试结果表明，所创建的随机森林回归模型预测值与实测值的平均绝对误差、平均相对误差随着特征变量数的增加呈下降趋势，特征变量数的增加可在一定程度上提高模型的预测效果。针对同一组数据，与主成分回归分析法相比，随机森林回归模型平均相对误差降低了14.29%，预测效果更好，且原理更简单、调参更容易、计算速度更快，能够为矿井回采工作面瓦斯涌出量预测提供有力的理论支撑。
- 回采工作面 /
- 瓦斯涌出量预测 /
- 随机森林回归 /
- 袋外数据评估分数 /
- 特征变量重要性 /
- 特征变量数
Abstract:
The mining face is the main place for gas emission in mines. Accurately predicting the amount of gas emission from the mining face and proposing targeted prevention and control measures are of great significance for ensuring mine safety production. A prediction method for gas emission in mining face based on random forest regression algorithm has been proposed. Using the measured gas emission data from the working face as the original sample, the Bootstrap sampling method is used for random sampling. The out-of-bag (OOB) data assessment score oob_score is used as an evaluation indicator for the random forest regression model tuning parameter and importance of feature variables. The optimal parameters of the model and the percentage of importance of feature variables are calculated. The method ranks the importance proportion of each feature variable and conducts performance analysis of the random forest regression model according to the ranking. The results show that as the number of feature variables increases, the model performance does not show a regular change. When the number of feature variables is small, there may be overfitting. The test results show that the average absolute error and relative error between the predicted and measured values of the created random forest regression model decrease with the increase of the number of feature variables. The increase of the number of feature variables can improve the predictive performance of the model to a certain extent. Compared with the principal component regression analysis method, the random forest regression model reduces the average relative error by 14.29% for the same set of data, resulting in better prediction performance. The principle is simpler, parameter adjustment is easier, and the calculation speed is faster. The results can provide strong theoretical support for predicting gas emission in mining face.
- mining face /
- gas emission prediction /
- random forest regression /
- out of bag data assessment score /
- importance of feature variables /
- number of feature variables

HTML全文

图 1 随机森林回归算法原理

Figure 1. Principle of random forest regression algorithm

下载: 全尺寸图片幻灯片

图 2 随机森林回归模型参数关系热力图

Figure 2. Thermodynamic diagram of random forest regression model parameter relationship

下载: 全尺寸图片幻灯片

图 3 n_estimators与obb_score的关系曲线

Figure 3. The relationship curves of n_estimators and obb_score

下载: 全尺寸图片幻灯片

图 4 特征变量重要性占比

Figure 4. The proportion of importance of characteristic variables

下载: 全尺寸图片幻灯片

图 5 特征变量数与oob_score关系

Figure 5. The relationship between the number of characteristic variables and oob_score

下载: 全尺寸图片幻灯片

图 6 特征变量数与误差平均值关系曲线

Figure 6. The relationship between the number of characteristic variables and mean value of error

下载: 全尺寸图片幻灯片

表 1 回采工作面瓦斯涌出量特征样本数据

Table 1. Sample data of gas emission characteristics in the mining face

序号	X₁/（m³·t⁻¹）	X₂/m	X₃/m	X₄/（°）	X₅/m	X₆/m	X₇/m	X₈	X₉/t	X₁₀	X₁₁/（m³·t⁻¹）	X₁₂/m	X₁₃/m	X₁₄	Y/（m³·min⁻¹）
1	1.92	408	2.0	10	2.0	4.42	155	0.96	1825	1	2.02	1.5	20	5.03	3.34
2	2.15	411	2.0	8	2.0	4.16	140	0.95	1527	1	2.1	1.21	22	4.87	2.97
3	2.14	420	1.8	11	1.8	4.13	175	0.95	1751	1	2.64	1.62	19	4.75	3.56
4	2.58	432	2.3	10	2.3	4.67	145	0.95	2078	1	2.4	1.48	17	4.91	3.62
5	2.40	456	2.2	15	2.2	4.51	160	0.94	2104	1	2.55	1.75	20	4.63	4.17
6	3.22	516	2.8	13	2.8	3.45	180	0.93	2242	1	2.21	1.72	12	4.78	4.6
7	2.80	527	2.5	17	2.5	3.28	180	0.94	1979	1	2.81	1.81	11	4.51	4.92
8	3.35	531	2.9	9	2.9	3.68	165	0.93	2288	1	1.88	1.42	13	4.82	4.78
9	3.61	550	2.9	12	2.9	4.02	155	0.92	2352	1	2.12	1.6	14	4.83	5.23
10	3.68	563	3.0	11	3.0	3.53	175	0.94	2410	1	3.11	1.46	12	4.53	5.56

下载: 导出CSV

表 2 随机森林回归模型主要待调参数

Table 2. The main parameters to be adjusted in the random forest regression model

序号	参数	参数说明
1	n_estimators	随机森林中决策树的数量
2	criterion	回归树衡量回归质量的指标
3	random_state	生成的森林模式
4	max_features	最佳分支时的特征个数
5	max_depth	决策树剪枝参数，防止模型过拟合。本次原始样本数量较少，不进行剪枝
	Min_sample_leaf
	Min_sample_spit
	Min_impurity_decrease

下载: 导出CSV

表 3 随机森林回归模型调参结果

Table 3. Parameter adjustment results of random forest regression model

criterion	n_estimators	max_ features	random_state	最大obb_score
mse	23	11	165	0.91575566
mae	20	14	70	0.92116429
friedman_mse	34	14	34	0.91395423

下载: 导出CSV

表 4 随机森林回归模型预测误差

Table 4. Prediction error of random forest regression model

特征变量个数	平均绝对误差/（m³·min⁻¹）	平均相对误差/%
3	0.228	5.030
4	0.174	3.72
5	0.109	3.21
6	0.091	2.36
7	0.090	2.32
8	0.147	3.38
9	0.124	3.05
10	0.083	2.15
11	0.053	1.72
12	0.109	2.73
13	0.026	1.28
14	0.005	0.77

下载: 导出CSV

表 5 不同预测模型预测结果对比

Table 5. Comparison of prediction results of different prediction models

实测值/ （m³·min⁻¹）	随机森林回归模型			主成分回归分析法
实测值/ （m³·min⁻¹）	预测值/ （m³·min⁻¹）	绝对误差/ （m³·min⁻¹）	相对误差/%	预测值/ （m³·min⁻¹）	绝对误差/ （m³·min⁻¹）	相对误差/%
4.06	3.85	−0.21	5.17	4.01	−0.05	1.23
4.92	4.84	−0.08	1.63	5.30	0.38	7.72
8.04	7.56	−0.48	5.97	7.56	−0.48	5.97

下载: 导出CSV

参考文献(20)

[1]	AQ 1018−2006矿井瓦斯涌出量预测方法[S AQ 1018-2006 Prediction method of mine gas emission rate[S
[2]	国家煤矿安全监察局. 防治煤与瓦斯突出细则[M]. 北京:煤炭工业出版社,2019. State Administration of Coal Mine Safety. Rules of coal and gas outburst prevention[M]. Beijing:China Coal Industry Publishing House,2019.
[3]	国家安全生产监督管理总局. 煤矿安全规程[M]. 北京:煤炭工业出版社,2022. State Administration of Work Safety. Coal mine safety regulations[M]. Beijing:China Coal Industry Publishing House,2022.
[4]	秦玉金,苏伟伟,姜文忠,等. 我国矿井瓦斯涌出量预测技术研究进展及发展方向[J]. 煤矿安全,2020,51(10):52-59. QIN Yujin,SU Weiwei,JIANG Wenzhong,et al. Research progress and development direction of mine gas emission forecast technology in China[J]. Safety in Coal Mines,2020,51(10):52-59.
[5]	余永强,李伟雄,邱贤德,等. 用灰色系统理论预测平顶山一矿戊_8煤层瓦斯涌出[J]. 煤炭科学技术,2002,30(3):53-55,46. YU Yongqiang,LI Weixiong,QIU Xiande,et al. Application of gray system theory to predict gas emission from Wu_8 seam in No.1 Mine of Pingdingshan[J]. Coal Science and Technology,2002,30(3):53-55,46.
[6]	伍爱友,田云丽,宋译,等. 灰色系统理论在矿井瓦斯涌出量预测中的应用[J]. 煤炭学报,2005,30(5):47-50. WU Aiyou,TIAN Yunli,SONG Yi,et al. Application of the grey system theory for predicting the amount of mine gas emission in coal mine[J]. Journal of China Coal Society,2005,30(5):47-50.
[7]	徐刚,王磊,金洪伟,等. 因子分析法与BP神经网络耦合模型对回采工作面瓦斯涌出量预测[J]. 西安科技大学学报,2019,39(6):965-971. XU Gang,WANG Lei,JIN Hongwei,et al. Gas emission prediction in mining face by factor analysis and BP neural network coupling model[J]. Journal of Xi'an University of Science and Technology,2019,39(6):965-971.
[8]	徐威,何保,张振文. 基于灰色遗传神经网络法的矿井瓦斯涌出量预测[J]. 安全与环境学报,2011,11(2):176-178. XU Wei,HE Bao,ZHANG Zhenwen. Genetic neural network method based on gray mine gas emission prediction[J]. Journal of Safety and Environment,2011,11(2):176-178.
[9]	赵建会,孙榕鸿. 矿井回采工作面瓦斯涌出量预测新途径[J]. 西安科技大学学报,2011,31(6):708-710. ZHAO Jianhui,SUN Ronghong. New method of prediction for mine coalface gas emission[J]. Journal of Xi'an University of Science and Technology,2011,31(6):708-710.
[10]	熊祖强,王晓蕾. 矿井相对瓦斯涌出量动态无偏灰色马尔科夫预测[J]. 安全与环境学报,2015,15(3):15-18. XIONG Zuqiang,WANG Xiaolei. Application of the Markov bias-free dynamic grey model for the relative gas gush-out rate predict[J]. Journal of Safety and Environment,2015,15(3):15-18.
[11]	李树刚,马彦阳,林海飞,等. 基于因子分析法的瓦斯涌出量预测指标选取[J]. 西安科技大学学报,2017,37(4):461-466. LI Shugang,MA Yanyang,LIN Haifei,et al. Selection of gas emission prediction index based on factor analysis[J]. Journal of Xi'an University of Science and Technology,2017,37(4):461-466.
[12]	付华,谢森,徐耀松,等. 基于ACC−ENN算法的煤矿瓦斯涌出量动态预测模型研究[J]. 煤炭学报,2014,39(7):1296-1301. FU Hua,XIE Sen,XU Yaosong,et al. Gas emission dynamic prediction model of coal mine based on ACC-ENN algorithm[J]. Journal of China Coal Society,2014,39(7):1296-1301.
[13]	邓军,雷昌奎,曹凯,等. 采空区煤自燃预测的随机森林方法[J]. 煤炭学报,2018,43(10):2800-2808. DENG Jun,LEI Changkui,CAO Kai,et al. Random forest method for predicting coal spontaneous combustion in gob[J]. Journal of China Coal Society,2018,43(10):2800-2808.
[14]	汪明,王建军. 基于随机森林的回采工作面瓦斯涌出量预测模型[J]. 煤矿安全,2012,43(8):182-185. WANG Ming,WANG Jianjun. Gas emission prediction model of stope based on random forests[J]. Safety in Coal Mines,2012,43(8):182-185.
[15]	吴奉亮,霍源,高佳南. 基于随机森林回归的煤矿瓦斯涌出量预测方法[J]. 工矿自动化,2021,47(8):102-107. WU Fengliang,HUO Yuan,GAO Jianan. Coal mine gas emission prediction method based on random forest regression[J]. Industry and Mine Automation,2021,47(8):102-107.
[16]	成小雨,周爱桃,郭焱振,等. 基于随机森林与支持向量机的回采工作面瓦斯涌出量预测方法[J]. 煤矿安全,2022,53(10):205-211. CHENG Xiaoyu,ZHOU Aitao,GUO Yanzhen,et al. Prediction method of gas emission based on random forest and support vector machine[J]. Safety in Coal Mines,2022,53(10):205-211.
[17]	BREIMAN L. Random forests[J]. Machine learning,2001,45(1):5-32. doi: 10.1023/A:1010933404324
[18]	胡坤,王素珍,韩盛,等. 基于TLBO−LOIRE的回采工作面瓦斯涌出量预测[J]. 应用基础与工程科学学报,2017,25(5):1048-1056. HU Kun,WANG Suzhen,HAN Sheng,et al. Gas emission quantity prediction of working face based on TLBO-LOIRE method[J]. Journal of Basic Science and Engineering,2017,25(5):1048-1056.
[19]	姜杉杉. 基于LLE−WPA−BP算法的瓦斯涌出量预测研究[D]. 阜新:辽宁工程技术大学,2022. JIANG Shanshan. Study of gas emission prediction based on LLE-WPA-BP algorithm[D]. Fuxin:Liaoning Technical University,2022.
[20]	吕伏,梁冰,孙维吉,等. 基于主成分回归分析法的回采工作面瓦斯涌出量预测[J]. 煤炭学报,2012,37(1):113-116. LYU Fu,LIANG Bing,SUN Weiji,et al. Gas emission quantity prediction of working face based on principal component regression analysis method[J]. Journal of China Coal Society,2012,37(1):113-116.