TBM施工隧洞不良地质实时超前探测与掘进智能决策理论研究

In the construction of hydraulic tunnels or traffic tunnels, China has topped the world for the growing demand for tunnel boring machines (TBMs). However, there is no real-time geo-hazards ahead prospecting method, at home and abroad, that is suitable for the complex environment and fast excavation of TBM. The determination of excavation parameter usually depends on experience, which always lead to inefficient tunneling, water/mud inrush, tunnel collapses, TBM blocked or even damaged and casualties. Therefore, a further research on real-time geo-hazards ahead prospecting and excavation intelligent decision-making theory in TBM tunneling is in need. By using the rock vibration of TBM tunneling as seismic source, a breakthrough of the theory in high-resolution imaging and water-body recognition, will help to overcome the major difficulties in the real-time prospecting of geo-hazards like faults, fractured zones or water-bearing structures ahead of TBM. For the rock mass structure near TBM (0-15 m ahead), a precise modeling technique and a high-order multi-scale algorithm for micro-nano scale structures, will be studied for a quantitative prediction of the physical and mechanics parameters. Based on real-time ahead prospecting, the corresponding response of TBM tunneling parameters in different geological condition will be revealed to construct an intelligent decision-making theory of TBM tunneling parameters, especially when encountering adverse geology bodies. In the end, the basic theory, supporting software, and practical method of real-time geo-hazards ahead prospecting and excavation intelligent decision-making in TBM tunneling will be established. This project will fill the foreign and domestic blank and provide vital support for a fast and safe TBM tunnel excavation. It will also improve China’s core competence in TBM manufacturing and tunneling all over the world.

我国水工和交通隧洞建设中岩石掘进机(TBM)的需求增长迅速，居世界首位。然而，国内外没有适用于TBM复杂环境和快速掘进的不良地质实时超前探测方法，其掘进参数选择依靠经验，往往导致效率低下、突水突泥、塌方卡机甚至机毁人亡等严重后果。因此本项目开展TBM不良地质实时超前探测与智能决策理论研究。巧妙利用TBM破岩震动为震源，突破高精度成像、水体识别等理论瓶颈，破解断层破碎带、含水构造等不良地质实时探测的重大难题。研究TBM近前方(0-15m)岩体构造精细建模方法与细-微观结构高阶多尺度算法，实现其物理力学参数的定量预测。以实时超前探测为支撑，揭示不同地质和岩体条件下TBM掘进参数响应规律，构建掘进参数和不良地质应对方案的智能决策方法。最终形成TBM不良地质实时超前探测与智能决策的基础理论、支撑软件和实用方法，填补国内外空白，为保障TBM安全高效掘进、提升TBM施工和制造的核心竞争力提供关键支撑。

我国水工和交通隧洞建设中岩石掘进机(TBM)的需求增长迅速，居世界首位。然而，国内外没有适用于TBM复杂环境和快速掘进的不良地质实时超前探测方法，其掘进参数选择依靠经验，往往导致效率低下、突水突泥、塌方卡机甚至机毁人亡等严重后果。为此，本项目开展了TBM不良地质实时超前探测与智能决策理论研究。巧妙利用TBM破岩震动为震源，深入研究破岩地震波场特征，建立了“震源特性先导传感+波场信息阵列测量”的观测方式，创新了波场特征恢复和有效信息提取方法，解决了破岩震源科学利用的难题；进而研究了波速准确反演和成像方法，自主研发了专用仪器与软件，实现了断层等不良地质随TBM掘进实时探测、跟踪校正和动态成像。针对掘进面近前方岩体构造、物理力学参数预测难题，发展了基于红外深度成像的岩体结构精细扫描和智能识别方法，创新了跨孔地质雷达岩体内部构造精细刻画方法，实现了多元信息融合的岩体结构建模；基于建立的岩体结构模型，提出了岩体力学参数高阶多尺度预测模型，实现了近前方岩体抗压强度、抗拉强度等力学参数的快速动态预测。针对TBM安全高效掘进难题，研发了关键岩体参数原位感知技术与TBM岩-机大数据平台，提出了物理规律与数据挖掘双驱动的TBM岩-机关系建模方法；以此为基础，创新提出了“以不良地质防控和TBM额定能力为安全约束，以高效、低消耗为目标，以双目标动态优化为途径”的TBM主控掘进参数智能寻优控制方法，推动了TBM掘进由“依赖经验”向“准确感知、智能决策”的进步。最终形成了一套TBM施工不良地质实时超前探测和掘进参数智能决策的基础理论、关键技术和支撑软件。相关成果在新疆某工程TBM隧洞、大瑞铁路高黎贡山隧道TBM、杭州市第二水源输水通道TBM、珠江三角洲水资源配置工程TBM等成功应用，为保障TBM安全高效掘进、提升TBM施工和制造的核心竞争力提供了关键支撑。