非线性灰色-时序组合模型在隧道变形监测中的应用研究  

Research on the Application of the Nonlinear Combination Model of Grey Model and Time Series Analysis for Monitoring of Tunnel Deformation

作　　者：王豪威 杨海成[1] 张新霞[1] 

机构地区：[1]核工业航测遥感中心,石家庄050002

出　　处：《江西科学》2015年第4期561-565,共5页Jiangxi Science

摘　　要：隧道变形监测时由于受外界因素或系统因素的干扰,所获得的观测数据往往会含有误差或粗差,必须在建模之前对观测数据进行预处理。选用格拉布斯准则对观测数据进行粗差检验,可以有效地避免粗差对模型预测精度的干扰进而提高模型的预测精度,且将灰色时序组合模型用于隧道变形监测分析预报中,并与单一的GM(1,1)模型和时间序列模型的预测精度作对比,结果表明:组合模型在隧道变形预测分析预报中有较高的精度。Observation data in the tunnel deformation will generally contain gross error because of the interference of external factors and system factors,so we must be observation data preprocessing before building the mathematical model. In paper,observation data is tested by Rugby Booth,can effectively avoid interference gross error of prediction accuracy of the model and to improve the prediction accuracy of the model. This article will GM（ 1,1） model and ARMA（ p,q） model for analysis and forecasting deformation data of tunnel. Result of analysis is compared with another result by GM（ 1,1） and ARMA（ p,q） forecasting. Final validation of nonlinear combined model would be better in forecasting of tunnel.

关 键 词：隧道变形监测 粗差检验 格拉布斯准则 精度分析 

分 类 号：TU463[建筑科学—结构工程]