基于深度学习及数值模拟耦合的覆岩采动效应研究  

Study on Mining Effect of Overlying Strata Based on Coupling of Deep Learning and Numerical Simulation

作　　者：姚福艳 吕昌兴 杨晓 Yao Fuyan;Lü Changxing;Yang Xiao(School of Geoscience and Engineering,Shandong University of Science and Technology,Shandong Qingdao 266590;Huafeng Coal Mine,Xinwen Mining Industry Group Co.,Ltd.,Shandong Tai'an 271000;Shandong Zhengyuan Construction Engineering Co.,Ltd.,Shandong Tai'an 271000)

机构地区：[1]山东科技大学地球科学与工程学院,山东青岛266590 [2]新汶矿业集团有限责任公司华丰煤矿,山东泰安271000 [3]山东正元建设工程有限责任公司,山东泰安271000

出　　处：《山东煤炭科技》2022年第8期180-184,188,共6页Shandong Coal Science and Technology

摘　　要：为研究回采工作面推进过程中覆岩裂隙发育规律进而论证煤柱留设尺寸的合理性,本文采用数理分析手段进行分析,将待开采工作面实测数据代入神经网络,获得了导水裂隙带预测值;借助FLAC^(3D)软件模拟工作面回采过程,得到了采场塑性区分布范围图。综合导水带预测值、塑性区分布图与断层实测产状验证,说明了煤柱留设的合理性。In order to study the development law of overlying rock fissures in the advancing process of mining working face and demonstrate the rationality of setting the size of coal pillar,this paper uses the means of mathematical analysis,substitutes the measured data of mining working face into neural network,obtains the predicted value of water diversion fracture zone,and obtains the distribution range map of stope plastic zone with the help of FLAC^(3D) software to simulate the mining process of working face.The comprehensive verification of the predicted value of the water diversion zone,the distribution map of the plastic zone and the measured occurrence of the fault shows that the setting of the coal pillar is reasonable.

关 键 词：矿井地质 导水裂隙带 BP神经网络 FLAC^(3D) 

分 类 号：TD745[矿业工程—矿井通风与安全]