煤体应力随钻监测响应特征及智能感知技术研究  

Research on drilling characteristics and the intelligent perception technology of coal stress

作　　者：王桂峰[1] 张崟峰 王忠宾[2] 静涛 窦林名[1] 曹安业[1] 胡非凡 刘心强 冯淦 WANG Guifeng;ZHANG Yinfeng;WANG Zhongbin;JING Tao;DOU Linming;CAO Anye;HU Feifan;LIU Xinqiang;FENG Gan(State Key Laboratory for Fine Exploration and Intelligent Development of Coal Resources,School of mines,China University of Mining and Technology,Xuzhou,Jiangsu 221116,China;School of Mechanical and Electrical Engineering,China University of Mining and Technology,Xuzhou,Jiangsu 221116,China;China Mining Dingbei Technology(Xuzhou)Co Ltd,Xuzhou,Jiangsu 221116,China;College of Water Resource&Hydropower,Sichuan University,Chengdu,Sichuan 610065,China)

机构地区：[1]中国矿业大学煤炭精细勘探与智能开发全国重点实验室,矿业工程学院,江苏徐州221116 [2]中国矿业大学机电学院,江苏徐州221116 [3]中矿鼎北科技(徐州)有限公司,江苏徐州221116 [4]四川大学水利水电学院,四川成都610065

出　　处：《采矿与安全工程学报》2024年第5期1069-1077,共9页Journal of Mining & Safety Engineering

基　　金：国家重点研发计划项目(2020YFB1314203,2022YFC3004603);国家重大科研仪器研制项目(52227901);国家自然科学基金项目(51504248,52227901,51934007);江苏省自然科学基金项目(BK20140194);江苏高校优势学科建设工程项目(SZBF2011-6-B35);中国矿业大学大学生创新创业训练计划项目(202210290422E)。

摘　　要：由于冲击地压灾害的复杂性、突发性及形式多样性等特点,灾源准确感知与预警、应力场精准调控等问题迫切需要解决,数字钻探被认为是定量测量岩体力学信息的原位测试新方法。通过自主研制的三轴高应力加载钻进全过程模拟实验系统,开展不同应力、钻进工况条件下钻进敏感性试验,研究不同应力、钻进工况对随钻扭矩、推力、排屑量等信息的影响规律,建立煤体应力随钻感知力学模型,揭示煤体应力与随钻响应信息关联机制,提出基于BP神经网络的煤体应力多参量融合智能识别方法,并开展了工程应用。研究表明:随钻响应信息(扭矩、推力等)与加载应力呈正相关关系,其增长趋势与钻进排屑量随应力增加的趋势相同,而钻进速度、转速是影响随钻响应信息的重要因素;基于BP神经网络的随钻应力识别模型拟合程度好,模型预测准确率达到95%;井下工程验证发现主动CT反演的应力异常区和随钻反演应力具有较好的一致性,验证了随钻感知煤体应力的准确性。研究为实现煤体应力原位数字钻探技术提供了科学依据,有利于破解冲击地压危险源智能感知与预警行业性难题,促进防冲技术向高效、智能化方向发展。Due to the complexity,suddenness,and diversity of forms of rockburst disasters,issues like accurate perception and warning of disaster sources and precise control of stress fields urgently need to be addressed.Digital drilling is considered a new in⁃situ testing method for quantitatively measuring rock mechanics information.An experimental system with triaxial high⁃stress loading for drilling simulation was innovatively developed.With the aid of this experimental system,drilling sensitivity experiments were conducted under different stress and drilling conditions to investigate the influences of different con⁃ditions on various drilling information,including torque,thrust,and drill cuttings.Additionally,a me⁃chanical model for identifying coal seam stress while drilling was established,and the mechanism of cor⁃relation between coal seam stress and drilling response information was revealed.On this basis,a multi⁃parameter⁃fused intelligent identification method for coal seam stress based on back propagation neural network(BPNN)was proposed and applied to engineering practice.The following findings were ob⁃tained.Drilling information(torque,thrust,et al)is positively correlated with loading stress,and its growth trend is identical to the growth trend of drill cuttings with stress.Drilling speed and rotational speed are important factors affecting drilling response information.The BPNN⁃based identification model for coal seam stress while drilling achieves an excellent fitting effect,its prediction accuracy being as high as 95%.According to the application to underground engineering practice,the stress⁃anomalous area obtained through active computed tomography(CT)is highly consistent with the stress obtained by inversion while drilling,which verifies the accuracy of identifying coal seam stress while drilling.This study,which provides a scientific basis for the implementation of in⁃situ digital drilling technology for coal seam stress,is beneficial for achieving intelligent perception and early

关 键 词：深部开采 冲击地压 应力随钻监测 三轴试验 神经网络 

分 类 号：P631[天文地球—地质矿产勘探]