多元时间序列分析的滑坡演化阶段划分

发布时间：2018-04-15 03:07

本文选题：滑坡演化 + 多元时间序列　；参考：《地球科学》2016年04期

【摘要】：滑坡的时间-位移曲线一般具有3个阶段特征,即初始变形阶段、等速变形阶段和加速变形阶段,不同演化阶段加速度具有不同的变化特点.目前往往是依据对加速度曲线特征的分析来人为划分演化阶段,缺少相应的理论支持和定量计算.针对上述问题,选取月降雨量、月库水位高程变化量对滑坡的累计位移建立多因素的时间序列预测模型.然后利用Chow分割点检验理论,以所建模型中F和LR统计量最大值点作为分割点对滑坡演化阶段进行划分.以新滩滑坡和三峡库区白水河滑坡为例,利用累计位移、降雨及库水位变化数据进行计算验证.结果表明,对多元时间序列模型进行Chow分割点检验可对滑坡的演化阶段进行准确划分,为滑坡的临滑预警预报提供重要判据.
[Abstract]:The time-displacement curve of landslide generally has three stage characteristics: initial deformation stage, isokinetic deformation stage and accelerated deformation stage, and the acceleration of different evolution stages has different characteristics.At present, it is often based on the analysis of the characteristics of the acceleration curve to artificially divide the evolution stage, which lacks the corresponding theoretical support and quantitative calculation.In view of the above problems, a multi-factor time series prediction model is established by selecting the monthly rainfall and the variation of monthly reservoir water level elevation to the cumulative displacement of landslide.Then using the Chow partition point test theory, the maximum points of F and LR statistics in the established model are used as the segmentation points to divide the landslide evolution stage.Taking Xintan landslide and Baishuihe landslide in three Gorges Reservoir area as examples, the data of accumulative displacement, rainfall and reservoir water level change are used to calculate and verify.The results show that the multivariate time series model can be used to accurately divide the evolution stages of landslide by using Chow partition point test, which provides an important criterion for landslide prediction and early warning.
【作者单位】：中国地质大学数学与物理学院;
【基金】：国家重点基础研究发展计划(973计划)项目(No.2011CB710605) 国家自然科学基金重点项目(No.41230637)
【分类号】：P642.22

【相似文献】