滑坡孕育过程次声波特性与临滑前兆研究

Some landslides(especially burst-type landslide) have no obvious deformation signs before damage, thus deformation-based monitoring and early-warning method does not work very well for this type of landslide. However, the micro-failure generated by the gradual forming of the sliding surface in the landslide body is an objective physical phenomenon and can be detected in real-time by modern technique. This research program plans to detect and study the characteristics of infrasonic wave generated in the process of landslide formation, and to establish corresponding strategies for landslide monitoring and prevention. First, a sophisticated system for detecting and analyzing infrasonic wave generated by landslide is developed to carry out continuous and long-term monitoring on the infrasonic wave and related parameters produced by some deforming landslides. The detected signals are analyzed in time-domain and frequency-domain to reveal the dynamic evolution characteristics of infrasound during the process of deformation and failure of potential landslide. The abnormal infrasonic information and precursory of landslide are studied in depth in order to explore and establish an infrasound-based monitoring and early-warning method for landslide.

某些滑坡(尤其是突发型滑坡)往往在整体失稳破坏前并不会表现出明显的宏观变形迹象，因此，常用基于变形的监测预警方法在此类滑坡中并不太适用。但对于任何滑坡而言，因滑动面逐渐形成与贯通所造成的坡体内部破裂却是客观存在的，并可借助现代探测技术实时探测出来。本申请课题拟重点探测研究滑坡孕育过程中的次声波特性，建立相应的滑坡监测预警方法。首先，研制高性能滑坡次声波探测分析系统，对一些正在变形的典型滑坡开展长期持续次声及相关物理参数监测，利用优化动态信号处理与分析方法分析监测结果，揭示滑坡孕育过程（变形破坏过程）次声波时域、频域多维空间特征和动态演化特性。在此基础上，重点分析研究滑坡发生前的次声波异常信息和前兆特征，以望探索建立基于次声波的滑坡监测预警方法。

滑坡（尤其是突发型滑坡）在整体失稳破坏之前并未表现出明显的宏观变形现象，传统变形监测预警方法较难适用于该类滑坡。然而，对于任何滑坡灾变过程而言，滑动面逐渐形成和贯通所导致的岩土体内部损伤破裂却是客观存在的，并可借助现代高精度探测技术与仪器实现对岩土体破裂及运动过程的探测。本项目采用地质灾害理论、声学、仪器科学、信号学等多学科交叉的研究思路，主要开展次声波探测仪器与信号处理分析方法研究，自主研发了数字化次声波探测仪器，研发了基于MATLAB GUI的声波事件自动识别、提取、EMD分解、时频分析、特征提取的综合分析软件，形成了低频次声波探测分析的整套软硬件方法体系，实现了对岩石破裂及滑坡孕育过程发射低频声波信息的高性能探测、处理与分析；在野外建立现场观测试验，现场探测岩质崩滑灾害孕育及运动过程的低频次声波信号，经过深入分析研究，揭示了崩塌过程辐射次声波信号在时频域特征及规律；基于人工智能与机器学习，构建了滑坡灾变过程的超前预测方法和岩石破裂类型与声信号特征的非监督分类方法，揭示了外界诱发因素与滑坡灾变的非线性耦合关系，以及岩石破裂机制演化过程声学信号特征的关系，将在未来几年的研究中深入加强，结合监测数据驱动与人工智能算法解决突发型滑坡监测预警的技术瓶颈。本项目研究成果及研发的仪器设备方法能为突发型滑坡监测预警方法提供技术支撑。