土性参数空间变异性的定量分析方法及其在滑坡风险评估中的应用

There exists a need to exactly determine soil properties for rational analysis and prevention of slope failure for hydraulic and hydro-power structures. However, geotechnical materials are natural materials, and soil properties vary spatially due to various geological factors. Quantifying such spatial variability of soil properties is a significant prerequisite for exact determination of soil properties. Existing statistical methods generally require a large amount of data to quantify the spatial variability, which is rarely satisfied in practice. The project aims to develop several Bayesian approaches to quantify the spatial variability of soil properties. The approach utilizes the limited information, which is collected during geotechnical site characterization, to identify statistically homogenous soil layers and to determine the correlation functions and statistics of soil properties concerned. Using the proposed Bayesian approach, effects of geological factors on the correlation function of soil properties are discussed. Then, random field theory is applied to simulate the spatial variability of soil properties according to the outcomes provided by the Bayesian approach. Based on the random simulation, this project develops an approach to identify key slope slip surfaces with the consideration of spatial variability. Using the proposed approach, the project will further explore how to quantify the relative contributions of different slope slip surfaces to the landslide risk. The project will not only lead to a significant improvement of the approaches for the exact determination of soil properties and their spatial variability, but also boost the applications of slope reliability analysis methods and landslide risk assessment approaches for preventing slope failure in geotechnical practice.

准确地确定土性参数是合理分析水利水电工程滑坡灾害的重要前提。然而，岩土体是一种天然材料，受复杂地质成因影响，土性参数值随着空间位置的变化而不同。定量地描述土性参数空间变异性对准确地确定土性参数至关重要。采用既有统计分析方法定量描述空间变异性通常需要大量数据，难以满足。本项目将提出一套贝叶斯方法，根据有限的岩土工程勘察资料定量地描述土性参数空间变异性，确定土性参数的统计均匀土层空间划分情况（如土层厚度和数目）、相关函数以及统计特征，并探讨地质成因对土性参数相关函数的影响规律。在此基础上，采用随机场理论模拟土性参数空间变异性，提出考虑空间变异性条件下边坡关键滑动面的识别方法，建立不同滑动面对滑坡风险的相对贡献的定量评价方法。本项目的研究成果不仅将推进土性参数及其空间变异性的确定方法的进一步发展，而且将推动边坡可靠度分析与滑坡风险评估方法在滑坡灾害防治工程实践中的应用，具有广泛的工程应用前景。

岩土体是一种天然材料，受复杂地质成因影响，土性参数值随着空间位置的变化而不同。定量地描述土性参数空间变异性对合理地确定土性参数至关重要。采用既有统计分析方法定量描述土性参数空间变异性通常需要大量数据，然而工程实际中勘探数据十分有限是滑坡风险定量评估中的关键瓶颈问题之一。针对该问题，本项目开展了有限数据条件下土性参数空间变异性定量表征方法和考虑空间变异性条件下滑坡风险评估研究。提出了一套基于贝叶斯理论的岩土工程概率化勘察方法，充分利用有限勘察资料，定量地描述土性参数空间变异性，确定统计均匀土层的空间划分情况以及每一个土层内土性参数的相关函数和统计特征，探讨了土体类别对土性参数相关函数和统计特征的影响。在此基础上，提出了考虑空间变异性条件下边坡系统可靠度高效分析方法，建立了不同滑动面对滑坡风险的相对贡献的定量评价指标和关键滑动面的判别准则，并开发了岩土工程概率化勘察软件和边坡稳定可靠度分析与风险评估软件。. 在项目支持下，出版英文专著1部（Springer出版社），发表学术论文31篇，其中SCI论文13篇（中科院大类3区以上论文12篇，包括中科院大类1区Top和2区Top论文2篇，4篇论文入选ESI高被引论文），EI论文10篇，国际会议论文3篇；获得软件著作权3项、外观专利1项；培养博士毕业生2人，硕士毕业生2人。项目研究成果推动了岩土工程可靠度分析和风险评估方法在滑坡灾害防治工程实践中的应用，为我国边坡稳定可靠度评价和滑坡风险评估提供了技术支撑。