逐步回归分析在桥梁挠度变形预测中的应用  

The applications of stepwise regression analysis for prediction of bridge deflection

作　　者：沈月千[1] 黄腾[1] 陈喜凤[1] 尤文进[1] 

机构地区：[1]河海大学地球科学与工程学院,南京210098

出　　处：《工程勘察》2012年第7期79-83,共5页Geotechnical Investigation & Surveying

摘　　要：本文结合某长江特大桥挠度监测实例,介绍逐步回归分析预测模型的建立,着重探讨挠度变形的主要影响因素,借助MATLAB软件从非桥梁结构的角度预测该桥不同时期的挠度变形值,并与基于灰色关联分析的线性回归预测结果进行对比分析。结果表明,逐步回归分析预测模型具有较高的精度,有较高的可靠性,为桥梁的运行管理提供科学依据。In this paper, based on the deflection monitoring project in Yangtze bridge, the forecasting modeling for stepwise regression analysis is introduced; the main influence factors of deflection deformation is studied; a MATLAB software is use to forecast the deformation values of a bridge deflection at different time under the perspective of non-bridge structure, and compared to the predictions of gray relational analysis based on linear regression. The results show that the accuracy and reliability of stepwise regression analysis is higher, which provides the scientific basis for the bridge operation management.

关 键 词：逐步回归分析 挠度预测 对比分析 影响因素 

分 类 号：TV698.1[水利工程—水利水电工程]