考虑变形成因机制的顺层边坡稳定性预测模型研究

The research of prediction of landslide hazard is frontier topic in international landslide hazard and environmental geology. Based on the construction of major engineering projects, the frontier issues such as the deformation response mechanism and stability prediction method of bedding slope are researched in detail, through the laboratory test, theoretical research, mathematical modeling and intelligent computing. The physical model experiment simulating the excavation process of bedding slope will be carried out to reveal the deformation regularity and mechanism of bedding slope under excavation. Combining the results of model experiment and Mindlin solution, a analytical model of deformation response of bedding slope excavation will be established which can reflect the rebound effect of unloading and structural characteristics of rock mass. Via the investigation and analysis of of monitoring data of major engineering projects, the affecting factors and regularity of slope will be resolved and the contribution of different factors on slope deformation will be obtained. On this basis, forecasting model of bedding slope stability considering deformation mechanism will be established based on multi-objective fuzzy optimization and particle swarm optimization algorithm, which can reduce the fitting error and enhance the development of the correlation simultaneously. The model can improve the reliability of prediction and early warning for slope and explain slope deformation legitimately, and it will help to prevention and control the landslide hazard in major engineering projects, which has important theoretical and engineering application value.

开展滑坡灾害预测研究是当今国际滑坡灾害研究和环境地质研究领域的前沿课题。本课题以重大工程为依托，综合运用室内试验、力学建模、理论分析及智能计算方法，对顺层边坡变形响应机理和稳定性预测方法等前沿问题开展研究。采用地质力学模型试验技术开展顺层边坡开挖变形响应的物理模型研究，揭示开挖卸荷作用下顺层边坡的变形规律与机理；结合模型试验结果和Mindlin应变解，建立能够反映卸荷回弹效应和岩体结构特征的顺层边坡开挖变形响应的解析模型；通过对重大工程监测资料的调研与分析，解析除开挖以外的其他变形影响因素及其影响规律, 获得不同影响因素对边坡变形的贡献程度；在此基础上，以"拟合误差小"和"发展相关性好"为目标，结合多目标模糊优化算法和粒子群算法，建立考虑变形成因机制的边坡稳定性预测模型。可以提高预测的可靠性，并合理解释边坡变形产生的原因，对重大工程中滑坡灾害的预防与控制具有重要的理论意义和工程应用价值。

关于滑坡预测预报的研究目前在国际上仍然是一个难题。如何对滑坡进行预测预报是滑坡防灾减灾的关键环节，只有以足够的精度，成功地预测预报出滑坡的发生时间、规模和发展趋势等，才能真正达到防灾减灾的目的。项目为深入研究边/滑坡灾害的预测预警问题，综合运用模型试验、力学分析及智能预测方法，开展了顺层边坡开挖变形响应的地质力学模型试验、基于Mindlin应变解的顺层边坡开挖变形响应解析、基于多目标模糊优化算法的边坡变形组合预测、滑坡不同变形阶段演化规律与变形速率预警判据等方面的研究，主要取得了如下成果：.（1）开展不同开挖与锚固过程顺层边坡地质力学模型试验，分析了边坡不同部位对不同开挖坡率、锚固次序等施工过程的力学响应规律。揭示了开挖卸荷作用下顺层边坡的变形规律与机理，可以为顺层边坡施工期稳定性评价及安全预报提供科学依据。.（2）综合开挖作用下边坡卸荷回弹变形模型与基于Kalhaway本构的顺层边坡层面滑移计算模型，提出了基于Mindlin应变解的顺层边坡开挖变形响应解析模型，能够反映开挖卸荷作用机制和岩体结构特征，可以在力学机制方面为顺层边坡开挖变形机理提供理论依据。.（3）总结了影响顺层边坡开挖变形的主要因素，提出了考虑变形成因机制的边坡变形稳定性预测模型，能够体现各影响因素对边坡变形的作用和相关程度，对边坡变形产生和发展的原因做出合理的物理力学解释。.（4）为提高边坡变形时间曲线预测精度，提出了基于多目标模糊优化算法的边坡变形组合预测模型，在影响因素变化的情况下提高多步预测的精度，可以实现边坡变形的多目标精细化预测。.（5）基于49个滑坡的过程信息，总结了滑坡不同变形阶段历时、裂缝特征与位移速率临界值，可以为滑坡灾害的预警与处治方案决策提供重要的参考依据。