稳健回归分析在斜拉桥温度效应分析中的应用

在桥梁施工过程中,桥梁各构件温度变化对主梁应变的作用机理是复杂的,同时因现场所采数据精度受现场环境因素影响较大,以致在研究桥梁系统温度与桥梁各构件温差对主梁应变作用的规律时增加了难度。为掌握其规律以便预测箱梁的温度效应,基于稳健最小二乘估计理论,建立关于主梁应变与各构件温差的稳健回归模型,对其进行计算分析。结果表明,稳健回归分析具备抗差性,其相应模型拟合优度较好,与多元线性回归分析相比,稳健回归分析在桥梁应变监测应用中效果较好。

Application of a robust regression model in the analysis of temperature effect of cable-stayed bridges

In construction,bridges are usually affected by temperature.The temperature effect is very complex,because each type of temperature actions has its own mechanism and leads to different effects on bridge structures.It is difficult to find out how temperature of a bridge system and temperature difference of bridge components affect girder strain.In order to predict the temperature effect of box girder,based on robust estimation theory,a robust regression model about the relationship between girder strain and temperature difference of bridge components is established.The analysis results show that the robust regression analysis has good robustness,and the fitting is good.Compared with the multiple linear regression analysis,the robust regression analysis is better in monitoring bridge strain.