高边坡变形非线性时变统计模型研究  

Research on nonlinear time series evolution statistic model of high slope displacements

作　　者：谈小龙[1,2] 徐卫亚[1,2] 梁桂兰[1] 孟永东[3] 张金龙[1] 

机构地区：[1]河海大学岩土工程科学研究所,南京210098 [2]岩土力学与堤坝工程教育部重点实验室,南京210098 [3]三峡大学土木水电学院,湖北宜昌443002

出　　处：《岩土力学》2010年第5期1633-1637,1650,共6页Rock and Soil Mechanics

基　　金：国家自然科学基金重点项目(No.50539110);国家科技支撑计划(No.2008BAB29B01);国家科技支撑计划(No.2006BAB04A02)资助课题

摘　　要：基于高边坡长期变形监测资料,深入分析了边坡变形的影响因素,建立边坡非线形时变统计监控模型,并将模型应用于锦屏一级水电站边坡工程。通过典型测点长期监测资料的分析,确定时效和降雨是影响边坡变形的2个主要因素,并给出了各自的表述形式,在Matlab平台上,采用非线性回归方法得到边坡非线性时变统计模型的参数。在此基础上,根据边坡变形拟合残差时序,采用ARMA模型方法进行拟合和预测,对边坡非线性时变模型的拟合和预测进行修正。结果表明,模型能较好地描述高边坡的变形特征,并且具有很好的拟合和预测精度。Based on long term displacement monitoring data of high rock slope,the factors of slope displacement are presented firstly.Then the monitoring model for high rock slope,called nonlinear time series evolution statistic monitoring model is established.It is also applied to the high slope project of Jinping Hydropower Station.Time-effect and rainfall are assumed as two main factors of high slope displacement;and their formulas are also presented through the analysis of long term monitoring data.The parameters of the monitoring model are obtained by the nonlinear regression technique on the platform of Matlab.Meanwhile the monitoring model is improved by using ARMA method to fit and predict the residual of regression of slope displacement.The result shows that the monitoring model can characterize the behavior of high slope displacement and has a reliable fitting and predicting precision.

关 键 词：高边坡位移 非线性 时变统计模型 ARMA模型 

分 类 号：TU471[建筑科学—结构工程]