【作者】 夏彬伟；

【导师】 李晓红； 卢义玉；

【作者基本信息】 重庆大学 ， 采矿工程， 2006， 硕士

【摘要】 在现代隧道建设中，发展趋势是隧道越建越长，穿越的地层地质条件越来越复杂，由于隧道工程的复杂性和不可预见性，在现有的技术经济条件下，隧道施工中各种不良地质灾害的预测和治理是复杂条件下山岭公路隧道施工中面临的最主要任务，尤其是隧道施工开挖工作面前方地质情况的预报是国内外工程地质和隧道工程界关注而又没有得到很好解决的难题。实践证明，在隧道施工中开展灾害预测预报技术能极大的减少塌方、突水、突泥等不良地质灾害，既可保证施工的顺利进行，又能极大的降低成本，故其在隧道施工中非常重要且十分必要。 本文结合国家自然科学基金重点项目“隧道及地下空间工程结构物的稳定性与可靠性”并依托于分水及笔架山隧道和通渝隧道开展了对硬质岩中的塌方、隧道前方不良地质体(断裂、溶洞、破碎带)规模及位置的确定、岩爆等三个方面进行预测预报的研究，得到以下主要结论： (1) 在分水及笔架山隧道施工过程中，运用赤平投影、地下洞室围岩块体稳定性分析程序对硬质岩中的不稳定块体进行检索，得出了不稳定块体与隧道的相对位置和不稳定块体的安全系数，从而对不稳定块体引起的坍塌进行了预测，在施工过程中加强了支护措施，避免了不稳定块体的坍塌。 (2) 通过利用地质雷达和工程多波地震仪对分水及笔架山隧道的探测，及时预报了掌子面前方围岩结构的情况，探测结果与实际揭露的围岩情况基本吻合，在定性和定量预报上达到了很高准确性，从而为优化隧道施工方案提供了依据，为预防隧道可能的发生灾害性事故及时提供了可靠的信息，达到了安全、快速施工的目的。 (3) 本文利用Visual C++6.0开发了基于BP神经网络的岩爆预测系统，并选择隧道围岩最大切向应力、岩石单轴抗压强度、岩石单轴抗拉强度、岩石弹性能量指数、围岩最大切向应力与岩石抗压强度的比值σ_θ／σ_c和岩石抗压强度与抗拉强度的比值σ_c／σ_t这六个指标作为BP人工神经网络预测岩爆的输入层，对通渝隧道岩爆预测进行了预测，其预测结果中表明，该方法是可行、可靠的。 (4) 利用有限元数值模拟计算出隧道围岩应力，参照国内外的岩爆发生的判据，预测岩爆发生，其结果较为可靠，同时也验证了利用BP人工神经网络岩爆预测结果是可信的。 (5) 综合运用了地下洞室围岩块体稳定性分析程序、地质雷达、多波工程地震仪、BP神经网络和数值模拟等多种预测手段，针对隧道不同地质灾害进行了系更多还原

【Abstract】 The trend of modern tunnel is the increase of its length. Fast construction has become the main approach in the field of tunnel construction in this century. Due to the complexity and unpredictability of tunnel construction, under current technical and economic conditions, prediction and treatment of varieties of harmful geological disasters become the major task involved in mountain tunnel construction under complex conditions, especially predictions of varieties of geological disasters draw more and more concerns among the fields of engineering geology and tunnel construction home and abroad and is still not dissolved. It proved that prediction and forecasting during tunnel construction could dramatically decrease the occurrence of such geologic disasters as collapse, water and mud bursting, etc. It could guarantee the construction and decline the cost.Referring to research on the stability and reliabiity of tunnel and underground space structure(50334060), based on the study of three aspects including collapse in hard rock, determination of scale and position of harmful geological body (fault, karst cave, fracture zone) in front of tunnel, rockburst etc.in Fenshui and Bijiashan Tunnel,Tong-Yu Tunnel, this paper drew the following conclusions:(1) During Fenshui and bijiashan tunnel construction, apply software to refer to the unstable block in the hard rock, and achieve the relative position between unstable block and tunnel and safety coefficient of the unstable block, hence predict the collapse caused by the instability of the blocks.(2) Through the detection of geological radar and engineering multi-wave seismograph, it could detect abnormal situations , predict the position、 occurrence of the har(?)ful geological body and completeness of the surrounding rock and possibility of water containing in front of the tunnel face, so provide support for the construction units to optimize the construction plan, provide reliable and timely information to avoid possible tunnel accidents, achieving the goal of safe and fast construction.(3) This paper put forward to apply BP neural network to predict tunnel rock burst and developed rock burst prediction system based on Visual C++6.0, chose maximum shearing stress, rock axial compress strength, rock axial pull strength, rock elastic energy index, the ratio between maximum shearing stress and rock axial compress strength, and the ratio between rock axial compress strength and rock axial pull更多还原