Prediction of surrounding rock stability of coal roadway based on machine learning and its application
-
摘要: 煤巷围岩稳定性分类对指导现场岩体工程设计、施工、管理具有重要的理论和工程实际意义。本文选取了影响煤巷围岩稳定性的7个关键指标，采用现场案例、调查问卷和文献计量等方法收集样本并建立了围岩稳定性分类数据库，基于6种机器学习方法分别建立了煤巷围岩稳定性分类预测模型。经模型计算得出，神经网络和改进的支持向量机模型具有较高的预测准确性。将模型应用于霍州矿区实际工程，结果表明，神经网络和改进的支持向量机方法预测精度高、可靠性好。Abstract: The classification of surrounding rock stability of coal roadway has important theoretical and practical significance for the design, construction and management of on-site rock mass engineering.This paper selected seven key indexes that affect the surrounding rock stability of coal roadway, collected the samples through field cases collection, questionnaires and literature measurement, and established the surrounding rock stability classification database.By drawing on six machine learning methods, this study established the classification prediction models of surrounding rock stability of coal roadway accordingly.Through model calculation, it is concluded that the Neural Network and the improved Support Vector Machine model have higher prediction accuracy.The model is applied to the actual project of Huozhou mining area.Results show that the neural network and the improved support vector machine methods have high prediction accuracy and good reliability.
-
图 5 RBF-SVM模型在测试集上的预测结果
Figure 5. Prediction results of the RBF-SVM model on the test set
表 1 核函数种类
Table 1. Kernel function type
核函数名称 函数表达式 线性核函数 $K\left(x_i, x_j\right)=x_i^{\mathrm{T}} x_j $ 多项式核函数 $K\left(x_i, x_j\right)=\left[\gamma x_i^{\mathrm{T}} x_j+b\right]^d $ Sigmoid核函数 $K\left(x_i, x_j\right)=\tanh \left(\gamma x_i^{\mathrm{T}} x_j+b\right) $ 径向基核函数 $K\left(x_i, x_j\right)=\exp \left(-\gamma\left\|x_i-x_j\right\|^2\right) $ 
下载: 导出CSV
表 2 数据库初始样本
Table 2. Initial samples of database
序号 σt σb σc H L N X 类别 1 80.12 9.23 50.90 750 7 0.90 100 4 2 45.40 9.38 45.40 532 15 1.71 70 3 3 24.31 6.89 63.50 296 12 0.87 100 3 4 72.27 4.83 47.20 350 16 0.96 15 2 5 58.00 13.65 58.00 162 24 2.05 100 1 6 43.60 3.42 4.96 303 17 0.38 15 5 7 73.83 4.84 19.90 288 12 0.97 25 2 8 48.51 5.66 48.10 430 14 1.05 70 3 ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ 140 37.45 7.09 51.80 255 11 1.10 40 2
下载: 导出CSV
表 3 围岩分类数据归一化
Table 3. Normalization of surrounding rock classification data
序号 σt σb σc H L N X 类别 1 0.91 0.48 0.62 1.00 0.00 0.29 1.00 4 2 0.39 0.49 0.55 0.66 0.11 0.65 0.67 3 3 0.08 0.29 0.79 0.29 0.07 0.28 1.00 3 4 0.79 0.12 0.57 0.37 0.12 0.32 0.06 2 5 0.58 0.84 0.71 0.07 0.23 0.80 1.00 1 6 0.36 0.00 0.00 0.30 0.14 0.06 0.06 5 7 0.81 0.12 0.20 0.27 0.07 0.32 0.17 2 8 0.44 0.18 0.59 0.50 0.10 0.36 0.67 3 ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ 140 0.27 0.30 0.64 0.22 0.11 0.38 0.33 2
下载: 导出CSV
表 4 6种机器学习算法的精度
Table 4. Accuracy of 6 machine learning algorithms
算法 精度/% 算法 精度/% RBF-SVM 96.4 提升法 71.4 神经网络 92.9 随机森林 85.7 决策树 64.3 k均值聚类 60.7
下载: 导出CSV
表 5 工程应用巷道参数
Table 5. Roadway parameters for engineering application
序号 矿井名称 巷道名称 σt σb σc H L N X 类别 1 李雅庄矿 6031巷 27.5 9.38 65.1 639.0 8.0 1.5 20.0 3 2 辛置矿 2-608巷 44.8 5.70 48.1 536.0 14.0 1.7 70.0 4 3 曹村矿 11-11071巷 46.1 7.20 46.1 142.0 8.0 0.7 20.0 2 4 2-10311.2巷 78.6 8.60 57.9 500.0 8.0 1.8 80.0 2 5 三交河矿 2-6011巷 58.0 13.70 58.0 221.0 24.0 1.1 100.0 2 6 2-5131巷 58.0 13.70 58.0 206.0 24.0 2.1 20.0 2 7 2-5081巷 58.0 13.70 58.0 218.0 24.0 2.1 20.0 2 8 六采区轨道巷 58.0 13.70 58.0 162.0 24.0 2.1 100.0 2 9 2上-3162 62.7 10.40 42.5 280.0 80.0 4.0 100.0 1 10 木瓜矿 9-2031巷 44.4 3.40 5.0 288.0 17.5 0.2 20.0 5 11 9-2011巷 44.4 3.40 5.0 256.0 17.5 0.2 25.0 5 12 干河矿 二采区回风巷 65.1 4.10 21.4 523.0 12.0 1.6 100.0 3 13 2-1081巷道 58.3 5.30 49.9 535.0 12.0 0.4 30.0 3
下载: 导出CSV
-
[1] 侯朝炯, 王襄禹, 柏建彪, 等. 深部巷道围岩稳定性控制的基本理论与技术研究[J]. 中国矿业大学学报, 2021, 50(1): 1-12. doi: 10.3969/j.issn.1009-105x.2021.01.001Hou Chaojiong, Wang Xiangyu, Bai Jianbiao, et al. Basic theory and technology study of stability control for surrounding rock in deep roadway[J]. Journal of China University of Mining & Technology, 2021, 50(1): 1-12. doi: 10.3969/j.issn.1009-105x.2021.01.001 [2] 许传华, 任青文, 李瑞. 地下工程围岩稳定性分析方法研究进展[J]. 金属矿山, 2003(2): 34-37. doi: 10.3321/j.issn:1001-1250.2003.02.012Xu Chuanhua, Ren Qingwen, Li Rui. Advances in researching the stability analysis methods of the surrounding rock mass in underground engineering[J]. Metal Mine, 2003(2): 34-37. doi: 10.3321/j.issn:1001-1250.2003.02.012 [3] 康红普, 姜鹏飞, 高富强, 等. 掘进工作面围岩稳定性分析及快速成巷技术途径[J]. 煤炭学报, 2021, 46(7): 2023-2045. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB202107002.htmKang Hongpu, Jiang Pengfei, Gao Fuqiang, et al. Analysis on stability of rock surrounding heading faces and technical approaches for rapid heading[J]. Journal of China Coal Society, 2021, 46(7): 2023-2045. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB202107002.htm [4] 朱永建, 冯涛. 锚杆支护超长煤巷顶板稳定性动态分类研究[J]. 煤炭学报, 2012, 37(4): 565-570. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB201204007.htmZhu Yongjian, Feng Tao. Research on dynamic stability classification of bolt support for super-long coal roadway roof[J]. Journal of China Coal Society, 2012, 37(4): 565-570. https://www.cnki.com.cn/Article/CJFDTOTAL-MTXB201204007.htm [5] 杨小彬, 王逍遥, 周世禄, 等. 基于改进广义回归神经网络的工作面低氧预测模型研究[J]. 矿业科学学报, 2019, 4(5): 434-440. http://kykxxb.cumtb.edu.cn/article/id/243Yang Xiaobin, Wang Xiaoyao, Zhou Shilu, et al. Prediction model of working face hypoxia based on improved generalized regression neural network[J]. Journal of Mining Science and Technology, 2019, 4(5): 434-440. http://kykxxb.cumtb.edu.cn/article/id/243 [6] 赵学军, 李建. 一种基于深度学习的煤矸石检测方法[J]. 矿业科学学报, 2021, 6(6): 730-736. doi: 10.19606/j.cnki.jmst.2021.06.012Zhao Xuejun, Li Jian. A method of coal gangue detection based on deep learning[J]. Journal of Mining Science and Technology, 2021, 6(6): 730-736. doi: 10.19606/j.cnki.jmst.2021.06.012 [7] 赵洪宝, 刘瑞, 刘一洪, 等. 基于深度学习方法的矿山微震信号分类识别研究[J]. 矿业科学学报, 2022, 7(2): 166-174. doi: 10.19606/j.cnki.jmst.2022.02.003Zhao Hongbao, Liu Rui, Liu Yihong, et al. Research on classification and identification of mine microseismic signals based on deep learning method[J]. Journal of Mining Science and Technology, 2022, 7(2): 166-174. doi: 10.19606/j.cnki.jmst.2022.02.003 [8] 孔祥松, 单仁亮, 赵文峰, 等. 基于GA-BP的煤巷围岩稳定性分类与应用[J]. 煤矿安全, 2016, 47(5): 219-222. https://www.cnki.com.cn/Article/CJFDTOTAL-MKAQ201605058.htmKong Xiangsong, Shan Renliang, Zhao Wenfeng, et al. Surrounding rock stability classification and application of coal roadway based on GA-BP[J]. Safety in Coal Mines, 2016, 47(5): 219-222. https://www.cnki.com.cn/Article/CJFDTOTAL-MKAQ201605058.htm [9] 孙洁, 刘晓悦. 基于蚁群优化神经网络的巷道围岩稳定性预测[J]. 微电子学与计算机, 2008, 25(5): 117-119. https://www.cnki.com.cn/Article/CJFDTOTAL-WXYJ200805032.htmSun Jie, Liu Xiaoyue. Roadway's surrounding rocks stability predicting based on neural nnetwork using ant optimization algorithm[J]. Microelectronics & Computer, 2008, 25(5): 117-119. https://www.cnki.com.cn/Article/CJFDTOTAL-WXYJ200805032.htm [10] 常利营, 卢建移, 段波, 等. 基于Hopfield网络的地下工程围岩稳定性分类[J]. 岩土工程学报, 2011, 33(S1): 194-197. https://www.cnki.com.cn/Article/CJFDTOTAL-YTGC2011S1034.htmChang Liying, Lu Jianyi, Duan Bo, et al. Stability classification of adjoining rock of underground engineering based on Hopfield network[J]. Chinese Journal of Geotechnical Engineering, 2011, 33(S1): 194-197. https://www.cnki.com.cn/Article/CJFDTOTAL-YTGC2011S1034.htm [11] 朱一丁, 马文涛. 回采巷道围岩分类的支持向量机方法[J]. 采矿与安全工程学报, 2006, 23(3): 362-365. https://www.cnki.com.cn/Article/CJFDTOTAL-KSYL200603025.htmZhu Yiding, Ma Wentao. Method of support vector machines for classifying surrounding rocks of gateways[J]. Journal of Mining & Safety Engineering, 2006, 23(3): 362-365. https://www.cnki.com.cn/Article/CJFDTOTAL-KSYL200603025.htm [12] 赵汝星. 基于随机森林的回采巷道围岩稳定性分类[J]. 煤矿安全, 2014, 45(11): 200-202, 206. https://www.cnki.com.cn/Article/CJFDTOTAL-MKAQ201411058.htmZhao Ruxing. Classification of roadway surrounding rock stability based on random forest[J]. Safety in Coal Mines, 2014, 45(11): 200-202, 206. https://www.cnki.com.cn/Article/CJFDTOTAL-MKAQ201411058.htm [13] 徐飞, 徐卫亚, 温森, 等. 基于PSO-PP的围岩稳定性评价[J]. 岩土力学, 2010, 31(11): 3651-3655. https://www.cnki.com.cn/Article/CJFDTOTAL-YTLX201011049.htmXu Fei, Xu Weiya, Wen Sen, et al. Projection pursuit based on particle swarm optimization for evaluation of surrounding rock stability[J]. Rock and Soil Mechanics, 2010, 31(11): 3651-3655. https://www.cnki.com.cn/Article/CJFDTOTAL-YTLX201011049.htm [14] Zhu C X, Wang F G. Stability classification of surrounding rock based on support vector machine classification theory[C]// Communications and Information Processing, Aveiro, 2012. [15] 文海家, 黄健豪, 袁性涵, 等. 基于数值试验的山岭隧道围岩稳定性GIS-SVM预测[J]. 岩石力学与工程学报, 2020, 39(S1): 2920-2929. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2020S1033.htmWen Haijia, Huang Jianhao, Yuan Xinghan, et al. GIS-SVM prediction of surrounding rock stability in mountain tunnel based on numerical experiment[J]. Chinese Journal of Rock Mechanics and Engineering, 2020, 39(S1): 2920-2929. https://www.cnki.com.cn/Article/CJFDTOTAL-YSLX2020S1033.htm [16] Zheng X, Chen Q T, Li H, et al. Analysis of the surrounding rock's stability based on fuzzy comprehensive evaluation model[J]. Applied Mechanics and Materials, 2012, 226/227/228: 1390-1395. [17] 王茂盛, 朱晔, 李永亮, 等. 基于模糊聚类方法的煤巷围岩稳定性分类研究[J]. 矿业科学学报, 2018, 3(3): 238-245. http://kykxxb.cumtb.edu.cn/article/id/143Wang Maosheng, Zhu Ye, Li Yongliang, et al. Study on stability classification of surrounding rock of coal gateway based on fuzzy clustering method[J]. Journal of Mining Science and Technology, 2018, 3(3): 238-245. http://kykxxb.cumtb.edu.cn/article/id/143 [18] Mahdevari S, Shahriar K, Sharifzadeh M, et al. Stability prediction of gate roadways in longwall mining using artificial neural networks[J]. Neural Computing and Applications, 2017, 28(11): 3537-3555. [19] Deng G Z, Fu Y K. Fuzzy rule based classification method of surrounding rock stability of coal roadway using artificial intelligence algorithm[J]. Journal of Intelligent & Fuzzy Systems, 2021, 40(4): 8163-8171. [20] Zhang W, Li X B, Gong F Q. Stability classification model of mine-lane surrounding rock based on distance discriminant analysis method[J]. Journal of Central South University of Technology, 2008, 15(1): 117-120. [21] 许建华, 张学工, 李衍达. 支持向量机的新发展[J]. 控制与决策, 2004, 19(5): 481-484, 495. https://www.cnki.com.cn/Article/CJFDTOTAL-KZYC200405000.htmXu Jianhua, Zhang Xuegong, Li Yanda. Advances in support vector machines[J]. Control and Decision, 2004, 19(5): 481-484, 495. https://www.cnki.com.cn/Article/CJFDTOTAL-KZYC200405000.htm [22] 张学工. 关于统计学习理论与支持向量机[J]. 自动化学报, 2000, 26(1): 32-42. https://www.cnki.com.cn/Article/CJFDTOTAL-MOTO200001005.htmZhang Xuegong. Introduction to statistical learning theory and support vector machines[J]. Acta Automatica Sinica, 2000, 26(1): 32-42. https://www.cnki.com.cn/Article/CJFDTOTAL-MOTO200001005.htm [23] 李孝安, 康继昌, 蔡小斌, 等. 进化神经网络研究进展[J]. 控制与决策, 1998, 13(6): 617-623. https://www.cnki.com.cn/Article/CJFDTOTAL-KZYC806.000.htmLi Xiaoan, Kang Jichang, Cai Xiaobin, et al. A survey of evolutionary neural networks[J]. Control and Decision, 1998, 13(6): 617-623. https://www.cnki.com.cn/Article/CJFDTOTAL-KZYC806.000.htm [24] 黄海新, 吴迪, 文峰. 决策森林研究综述[J]. 电子技术应用, 2016, 42(12): 5-9. https://www.cnki.com.cn/Article/CJFDTOTAL-DZJY201612004.htmHuang Haixin, Wu Di, Wen Feng. Review of decision forest[J]. Application of Electronic Technique, 2016, 42(12): 5-9. https://www.cnki.com.cn/Article/CJFDTOTAL-DZJY201612004.htm [25] 姜明雪, 杨有龙. 基于密度峰值聚类和模糊支持度的boosting随机森林[J]. 南京大学学报: 自然科学, 2021, 57(4): 582-590. https://www.cnki.com.cn/Article/CJFDTOTAL-NJDZ202104006.htmJiang Mingxue, Yang Youlong. Density peak clustering and fuzzy support-based boosting random forest[J]. Journal of Nanjing University: Natural Science, 2021, 57(4): 582-590. https://www.cnki.com.cn/Article/CJFDTOTAL-NJDZ202104006.htm [26] 方匡南, 吴见彬, 朱建平, 等. 随机森林方法研究综述[J]. 统计与信息论坛, 2011, 26(3): 32-38. https://www.cnki.com.cn/Article/CJFDTOTAL-TJLT201103007.htmFang Kuangnan, Wu Jianbin, Zhu Jianping, et al. A review of technologies on random forests[J]. Statistics & Information Forum, 2011, 26(3): 32-38. https://www.cnki.com.cn/Article/CJFDTOTAL-TJLT201103007.htm [27] 杨俊闯, 赵超. K-Means聚类算法研究综述[J]. 计算机工程与应用, 2019, 55(23): 7-14, 63. https://www.cnki.com.cn/Article/CJFDTOTAL-JSGG201923003.htmYang Junchuang, Zhao Chao. Survey on K-means clustering algorithm[J]. Computer Engineering and Applications, 2019, 55(23): 7-14, 63. https://www.cnki.com.cn/Article/CJFDTOTAL-JSGG201923003.htm [28] 刘长平, 叶春明, 刘满成. 来自大自然的寻优策略: 像蝙蝠一样感知[J]. 计算机应用研究, 2013, 30(5): 1320-1322, 1356. https://www.cnki.com.cn/Article/CJFDTOTAL-JSYJ201305009.htmLiu Changping, Ye Chunming, Liu Mancheng. Optimization strategy from nature: perceive as bat[J]. Application Research of Computers, 2013, 30(5): 1320-1322, 1356. https://www.cnki.com.cn/Article/CJFDTOTAL-JSYJ201305009.htm [29] 孔祥松. 山西焦煤煤巷围岩稳定性分类与强帮强角支护技术研究[D]. 北京: 中国矿业大学(北京), 2014. [30] 王茂源. 煤巷锚杆支护设计混合智能系统研究[D]. 北京: 中国矿业大学(北京), 2016. [31] 杨仁树, 马鑫民. 煤矿巷道支护智能设计系统与工程应用[M]. 北京: 冶金工业出版社, 2015: 47-50. [32] 游皓麟. R语言预测实战[M]. 北京: 电子工业出版社, 2016. -
点击查看大图
图(12) / 表(5)
计量
- 文章访问数: 446
- HTML全文浏览量: 96
- PDF下载量: 63
- 被引次数: 0