主余震序列作用下滑坡灾害危险性概率评估方法研究

Earthquake-induced landslides are extremely hazardous, posing a severe threat to human lives and properties. Therefore, it is of great importance to assess the seismic landslide hazard in a rapid and accurate way, especially for seismically active regions. This project aims at conducting a probabilistic seismic slope displacement hazard analysis for slopes subjected to mainshock-aftershock earthquake sequences, by studying ground motion characteristics of the main-aftershock sequences, the influence of aftershock on computed slope displacements, and developing displacement hazard curves. A ground motion selection approach based on dual-conditional response spectra of main-aftershock earthquakes is firstly proposed. Based on the Newmark rigid-block and fully coupled stick-slip models, permanent sliding displacements are computed using ground motions of the main-aftershock sequences selected; simple empirical models for rigid and flexible slopes are then developed, respectively. Finally, a new analytical and computationally efficient framework to rigorously evaluate the slope displacement hazard subjected to main-aftershock earthquake sequences is developed, by stochastically simulating all sources of uncertainties in a fully probabilistic analysis. The results of this study will provide theoretical basis and technical support for realistic prediction of seismic landslide hazard areas and the mitigation of earthquake-induced landslide disasters.

地震诱发滑坡危害巨大，给人民群众的生命财产安全构成严重威胁。因此，对地震滑坡灾害危险性进行快速、准确评估对地震高发地区的防灾减灾具有重要意义。本项目针对边坡在主余震序列作用下的动力反应和滑动位移开展研究，通过统计分析和数值计算，重点研究主余震序列地震动特性、余震效应对边坡滑移的影响以及主余震作用下滑坡危险性概率评估方法。通过建立主余震地震动参数多维高斯模型，提出考虑主余震双条件反应谱的选波方法；深入研究自然和工程边坡在主余震作用下的滑移分布规律，建立考虑余震效应的边坡滑移预测模型；提出主余震作用下边坡滑移全概率分析方法，建立基于高效随机模拟的滑坡危险性快速评估方法。研究成果将为合理预测地震滑坡危险性分布、减轻我国地震滑坡灾害提供理论基础和技术支持。

地震诱发滑坡危害巨大，给人民群众的生命财产安全构成严重威胁，但是，边坡在考虑主余震序列作用下的滑坡灾害危险性研究较少。本项目基于地震概率危险性分析、随机数值模拟、机器学习算法等研究手段，采用理论分析与数值模拟相结合的研究方法，建立了能够表征主余震序列震级分布和地震动参数内在相关性的优化模型，提出了针对主余震序列地震波输入的选波方法；研究了典型土坡在主余震序列作用下的动力响应与损伤特征，分析了土坡在主余震序列作用下的滑动永久位移，阐明了土质边坡在主余震序列作用下的滑移分布特征，建立了主余震位移预测模型以便于工程应用；提出了考虑土体参数空间随机性的主余震滑坡概率危害性评估方法，引入了高效预测边坡屈服加速度的多重响应面模型，实现了对地震动不确定性、位移预测不确定性以及土体参数空间变异性的综合考虑。. 在项目支持下，发表期刊论文16篇，其中SCI收录论文11篇、EI收录论文1篇；发表国际会议论文1篇；申请公开国家发明专利4项；培养博士毕业生2人，硕士毕业生1人。项目研究成果为主余震序列作用下的边坡动力响应分析和滑坡危险性评估提供了重要技术支撑。