Intelligent identification of electromagnetic radiation signals induced by coal rock fractures using machine learning
-
摘要: 电磁辐射作为一种有效监测技术已应用于冲击地压、煤与瓦斯突出等煤岩动力灾害监测预警，但因电磁信号产生机制复杂，易受井下环境干扰（干扰信号）而影响灾害危险监测预警准确性。准确辨识煤岩破裂诱发的电磁辐射信号（有效信号）是该技术应用推广的关键。开展了煤岩单轴压缩电磁辐射监测实验，分析了电磁辐射有效信号和干扰信号时域、频域及分形特征差异性，分别利用线性判别法、支持向量机和集成学习法等机器学习算法建立了电磁辐射有效信号和干扰信号智能辨识模型，并对比分析了不同模型的识别精度。结果表明：分形盒维数、平均频率、计数和峰值频率特征对电磁辐射有效信号和干扰信号区分较明显，单一特征识别准确率均在70%以上；信号特征集和机器学习算法对有效信号和干扰信号识别准确率均有影响，基于全部特征集的集成学习法识别准确率最高，对2类信号的平均识别准确率为94.5%，能够满足电磁辐射监测预警应用需求。Abstract: Electromagnetic radiation (EMR) has proven to be an effective monitoring technology for coal rock dynamic disasters, including underground rock burst and coal and gas outbursts. However, the intricate generation mechanisms of electromagnetic signal, coupled with interference from underground environments, can compromise the accuracy of disaster monitoring and early warning systems. Accurately identifying EMR signals induced by coal rock fractures (effective signals) is essential for the widespread application of this technology. This study conducted monitoring experiments on electromagnetic radiation during uniaxial compression of coal rock, analyzing the time-domain, frequency-domain, and fractal characteristics of both valid and interference signals. Machine learning algorithms, such as linear discriminant analysis, support vector machines, and ensemble learning methods, were utilized to develop intelligent identification models for effective and interference signals. A comparative analysis of the recognition accuracy across different models was performed. The results demonstrated that characteristics like fractal box dimension, average frequency, count, and peak frequency effectively distinguished between valid and interference signals, with single-feature recognition accuracy surpassing 70%. Both the feature set and the choice of machine learning algorithm significantly influenced the identification accuracy of valid and interference signals. The ensemble learning method, leveraging the complete feature set, achieved the highest identification accuracy of 94.5% for both signal types, fulfilling the requirements for EMR monitoring and early warning applications.
-
图 1 煤岩加载电磁辐射监测实验系统
Figure 1. Electromagnetic radiation monitoring experimental system for coal rock loading
图 2 试样加载过程应力、电磁辐射能量、计数时序变化
Figure 2. Stress, electromagnetic radiation energy, and counting during sample loading
图 3 典型电磁辐射有效信号与干扰信号
Figure 3. Typical effective signals of electromagnetic radiationand interference signals
图 4 电磁辐射信号时域特征
Figure 4. Time domain characteristics of electromagnetic radiation signals
图 5 有效信号与干扰信号时域特征统计
Figure 5. Statistics of time-domain characteristics for effective signals and interference signals
图 6 有效信号与干扰信号频域特征统计
Figure 6. Statistics of frequency domain characteristics of effective signals and interference signals
图 7 有效信号与干扰信号分形特征统计
Figure 7. Statistics of fractal characteristics of effective signals and interference signals
图 10 不同特征集及机器学习算法的识别准确率对比
Figure 10. Comparison of recognition accuracy for different feature sets and machine learning algorithms
表 1 电磁辐射信号时域特征计算方法
Table 1. Calculation methods for time-domain characteristics of electromagnetic radiation signals
信号特征 计算方法 含义 计数 信号整个波动过程超过触发阈值的振荡次数 计数越大，说明信号波动越剧烈 能量 经放大器放大后电磁辐射信号电压对时间的积分（包络线下面积） 表示信号整体波动情况，能量/绝对能量越大，说明超过触发阈值的幅值越大，持续时间越长 绝对能量 经放大器放大前电磁辐射信号电压平方的时间积分除以10 kΩ 上升时间 信号开始时刻到最大幅值对应时刻的时间间隔 上升时间越短，说明信号起伏波动越剧烈 持续时间 信号有效成分波动开始时刻到结束时刻的时间间隔 持续时间越长，说明信号中超过触发阈值的数据点越多 最大幅值 信号波动幅值的最大值，单位为dB，与电压的换算关系：A=20lg（Vm/Vr）－g，A为最大幅值，Vm为要换算的电压，Vr为参考电压，规定1 μV对应0 dB，g为触发阈值 表示信号波动的最大范围，幅值越大，说明煤岩破裂或干扰引起的幅值波动范围越大 均方根 $ {\text{RMS}} = \sqrt {\dfrac{1}{T}\displaystyle \int_{{T_1}}^{{T_2}} {{{V^2(t)}}{\mathrm{d}}t} } $ 表示信号幅值偏离均值的程度。均方根越大，说明信号波动过程中偏离均值的程度越大 方差 $ {S^{\text{2}}} = \displaystyle \sum\limits_{{t=T_1}}^{{T_2}} {{{(V(t) - \bar V)}^2}/(n - 1)} $，$ \bar V $为信号开始时刻到结束时刻之间的平均幅值，n为持续时间内采样点总数 与均方根含义类似，方差越大，说明信号波动的离散程度越大 
下载: 导出CSV
-
[1] 王恩元,何学秋,聂百胜,等. 电磁辐射法预测煤与瓦斯突出原理[J]. 中国矿业大学学报,2000,29(3):225-229. doi: 10.3321/j.issn:1000-1964.2000.03.001WANG Enyuan,HE Xueqiu,NIE Baisheng,et al. Principle of predicting coal and gas outburst using electromagnetic emission[J]. Journal of China University of Mining & Technology,2000,29(3):225-229. doi: 10.3321/j.issn:1000-1964.2000.03.001 [2] 李红平,李鹏鹏,张强,等. 煤岩双剪摩擦滑动声电响应实验研究[J]. 煤矿安全,2023,54(3):169-176.LI Hongping,LI Pengpeng,ZHANG Qiang,et al. Experimental study on acoustic and electrical response of double shear friction sliding of coal and rock[J]. Safety in Coal Mines,2023,54(3):169-176. [3] 宋大钊,童永军,邱黎明,等. 花岗岩劈裂破坏电磁−震动有效信号重构与混沌特征[J]. 煤炭学报,2024,49(3):1375-1387.SONG Dazhao,TONG Yongjun,QIU Liming,et al. Effective signal reconstruction and chaotic characteristics of electro-seismic signal of granite splitting failure[J]. Journal of China Coal Society,2024,49(3):1375-1387. [4] 金佩剑,王恩元,宋大钊,等. 单轴循环加载煤岩电磁辐射规律实验研究[J]. 煤矿安全,2013,44(5):46-48.JIN Peijian,WANG Enyuan,SONG Dazhao,et al. Experimental study on coal rock electromagnetic radiation laws under uniaxial cyclic loading[J]. Safety in Coal Mines,2013,44(5):46-48. [5] 艾迪昊,李成武,赵越超,等. 煤体静载破坏微震、电磁辐射及裂纹扩展特征研究[J]. 岩土力学,2020,41(6):2043-2051.AI Dihao,LI Chengwu,ZHAO Yuechao,et al. Investigation on micro-seismic,electromagnetic radiation and crack propagation characteristics of coal under static loading[J]. Rock and Soil Mechanics,2020,41(6):2043-2051. [6] 王恩元,刘晓斐,何学秋,等. 煤岩动力灾害声电协同监测技术及预警应用[J]. 中国矿业大学学报,2018,47(5):942-948.WANG Enyuan,LIU Xiaofei,HE Xueqiu,et al. Acoustic emission and electromagnetic radiation synchronized monitoring technology and early-warning application for coal and rock dynamic disaster[J]. Journal of China University of Mining & Technology,2018,47(5):942-948. [7] 何学秋,孙晓磊,殷山,等. 岩石破坏过程磁场效应实验研究及其对地震预报的意义[J]. 地球物理学报,2023,66(11):4609-4624. doi: 10.6038/cjg2022Q0732HE Xueqiu,SUN Xiaolei,YIN Shan,et al. Experimental research on magnetic field variation in rock failure process and its significance for earthquake prediction[J]. Chinese Journal of Geophysics,2023,66(11):4609-4624. doi: 10.6038/cjg2022Q0732 [8] RABINOVITCH A,BAHAT D,FRID V. Similarity and dissimilarity of electromagnetic radiation from carbonate rocks under compression,drilling and blasting[J]. International Journal of Rock Mechanics and Mining Sciences,2002,39(1):125-129. doi: 10.1016/S1365-1609(02)00012-6 [9] ZHU Chenwei,NIE Baisheng. Spectrum and energy distribution characteristic of electromagnetic emission signals during fracture of coal[J]. Procedia Engineering,2011,26:1447-1455. doi: 10.1016/j.proeng.2011.11.2322 [10] QIU Liming,LI Zhonghui,WANG Enyuan,et al. Characteristics and precursor information of electromagnetic signals of mining-induced coal and gas outburst[J]. Journal of Loss Prevention in the Process Industries,2018,54:206-215. doi: 10.1016/j.jlp.2018.04.004 [11] LI Baolin,LI Zhonghui,WANG Enyuan,et al. Discrimination of different AE and EMR signals during excavation of coal roadway based on wavelet transform[J]. Minerals,2022,12(1). DOI: 10.3390/min120100637 [12] 姚精明,董文山,闫永业,等. 受载煤岩体电磁辐射动态多重分形特征[J]. 煤炭学报,2016,41(6):1429-1433.YAO Jingming,DONG Wenshan,YAN Yongye,et al. Multi-fractal characteristics of electromagnetic radiation with loaded coal[J]. Journal of China Coal Society,2016,41(6):1429-1433. [13] 胡少斌,王恩元,李忠辉,等. 受载煤体电磁辐射动态非线性特征[J]. 中国矿业大学学报,2014,43(3):380-387.HU Shaobin,WANG Enyuan,LI Zhonghui,et al. Nonlinear dynamic characteristics of electromagnetic radiation during loading coal[J]. Journal of China University of Mining & Technology,2014,43(3):380-387. [14] 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.ZHAO 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. [15] LI Baolin,LI Nan,WANG Enyuan,et al. Discriminant model of coal mining microseismic and blasting signals based on waveform characteristics[J]. Shock and Vibration,2017(8):113-125. [16] 尚雪义,李夕兵,彭康,等. 基于EMD−SVD的矿山微震与爆破信号特征提取及分类方法[J]. 岩土工程学报,2016,38(10):1849-1858. doi: 10.11779/CJGE201610014SHANG Xueyi,LI Xibing,PENG Kang,et al. Feature extraction and classification of mine microseism and blast based on EMD-SVD[J]. Chinese Journal of Geotechnical Engineering,2016,38(10):1849-1858. doi: 10.11779/CJGE201610014 [17] 董陇军,张义涵,孙道元,等. 花岗岩破裂的声发射阶段特征及裂纹不稳定扩展状态识别[J]. 岩石力学与工程学报,2022,41(1):120-131.DONG Longjun,ZHANG Yihan,SUN Daoyuan,et al. Stage characteristics of acoustic emission and identification of unstable crack state for granite fractures[J]. Chinese Journal of Rock Mechanics and Engineering,2022,41(1):120-131. [18] 曾鹏,纪洪广,孙利辉,等. 不同围压下岩石声发射不可逆性及其主破裂前特征信息试验研究[J]. 岩石力学与工程学报,2016,35(7):1333-1340.ZENG Peng,JI Hongguang,SUN Lihui,et al. Experimental study of characteristics of irreversibility and fracture precursors of acoustic emission in rock under different confining pressures[J]. Chinese Journal of Rock Mechanics and Engineering,2016,35(7):1333-1340. [19] 邱黎明. 煤巷掘进突出危险性的声电瓦斯监测预警研究[D]. 徐州:中国矿业大学,2018.QIU Liming. Study on acoustic-electrical gas monitoring and early warning of outburst danger in coal roadway excavation[D]. Xuzhou:China University of Mining and Technology,2018. [20] 赵聪聪,唐绍辉,覃敏,等. 矿震震源时空分布的分形特性与活动性预测——以新疆阿舍勒铜矿为例[J]. 岩石力学与工程学报,2019,38(增刊1):3036-3044.ZHAO Congcong,TANG Shaohui,QIN Min,et al. Fractal characteristics of spatiotemporal distribution and activity prediction based on mine earthquake—taking the Ashele copper mine in Xinjiang as an example[J]. Chinese Journal of Rock Mechanics and Engineering,2019,38(S1):3036-3044. [21] 钟明寿,龙源,谢全民,等. 基于分形盒维数和多重分形的爆破地震波信号分析[J]. 振动与冲击,2010,29(1):7-11,233. doi: 10.3969/j.issn.1000-3835.2010.01.002ZHONG Mingshou,LONG Yuan,XIE Quanmin,et al. Signal analysis for blasting seismic wave based on fractal box-dimension and multi-fractal[J]. Journal of Vibration and Shock,2010,29(1):7-11,233. doi: 10.3969/j.issn.1000-3835.2010.01.002 [22] 于冰冰,李清,赵桐德,等. 基于SSA−SVM的巷道顶板空顶沉降量预测模型[J]. 煤炭学报,2024,49(增刊1):57-71.YU Bingbing,LI Qing,ZHAO Tongde,et al. Prediction model of empty roof settlement of roadway roof based on SSA-SVM[J]. Journal of China Coal Society,2024,49(S1):57-71. -
点击查看大图
计量
- 文章访问数: 75
- HTML全文浏览量: 23
- PDF下载量: 3
- 被引次数: 0
