基于纵向数据与突变理论的边坡滑坡预测新方法及其应用  

New method for landslide forecast and its application based on longitudinal data and catastrophic theory

作　　者：王树仁[1] 张艳博[2] 

机构地区：[1]燕山大学建筑工程与力学学院,河北秦皇岛066004 [2]河北理工大学资源与环境学院,河北唐山063009

出　　处：《煤炭学报》2009年第5期640-644,共5页Journal of China Coal Society

基　　金：河北省自然科学基金资助项目(E2008000410);燕山大学博士基金资助项目(B245)

摘　　要：基于边坡位移监测数据满足纵向数据部分线性模型的假定,采用广义最小二乘法求得回归系数和回归函数,应用B样条法对坡表水平位移和观测时间进行回归分析,得到纵向数据部分线性模型函数.然后将该函数依泰勒级数展开并截取部分函数项,通过变量代换转换成尖点突变模型形式,由突变判据进行边坡滑移危险性识别及边坡滑坡预测.该新方法的提出丰富了滑坡预报理论,并得到了现场实测数据的验证.Under the assumptions that the monitoring data of slope displacement will satisfy for the partial linear model with longitudinal data, the regression function and regression parameter were obtained by applying the generalized least square method, and the partial linear model was derived after the regression analysis for the slope horizontal displacement and monitoring time by using B-spline method. The front several terms of the Taylor＇ s series were intercepted after the function of the partial linear model was expressed as Taylor＇ s series expansion. The cusp catastrophic model was eventually obtained by means of variable substitution, and the risk of landslide would be recognized with the aid of the catastrophic criterion. The new method of landslide forecast was verified by the monitoring data in the field.

关 键 词：纵向数据 突变理论 边坡工程 

分 类 号：TD824.7[矿业工程—煤矿开采]