大坝变形变系数回归建模

在变形分析建模中常采用回归分析建模,但普通的回归模型是一种静态的模型,当变形体结构或物理性质发生变化时,普通线性回归所建立的静态模型将不再适用。变系数回归模型是一种动态模型,有着更强的灵活性和适应性,因此,将变系数回归引入大坝变形分析建模中,采用局部线性估计的方法进行系数拟合。仿真和大坝变形建模实验表明,变系数回归得到的大坝变形模型优于普通的线性回归模型,预测精度更高。

Modeling Dam Deformation Using Varying Coefficient Regression

Regression analysis is often used in deformation analysis modeling，however，it is a static model. When the structure or physical properties of the monitoring object change，the static model will no longer be applicable. However，a varying coefficient regression model is a dynamic model and has greater flexibility and adaptability. So varying coefficient regression is introduced into dam deformation analysis. Local linear estimation is used to fit the varying coefficients. Both a simulation experiment and a real dam deformation monitoring data experiment show that a dam deformation model using varying coefficient regression is better than that using ordinary linear regression;and more precise when predicting deformation.