基于有限元法和逐步回归法的泵站位移统计模型构建方法

由于大部分的泵站地基不是岩基,通过传统的力学方法无法得到较为准确的泵站位移统计模型,计算较为困难。因此,基于大坝变形统计模型构建的理念,提出了基于有限元-逐步回归分析的泵站位移统计模型构建方法。该方法以逐步回归为基础,通过有限单元法来确定统计模型中与水压分量有关的各个因子,再根据监测资料来确定温度分量和时效分量里的各个因子,从而得到较为准确的泵站建筑物位移统计模型。经南水北调东线某泵站验证表明,泵站实测的水平、垂直位移与三维有限元-逐步回归分析得出的结果基本一致,进一步表明,采用的统计模型用于泵站的位移分析是可行的。基于有限元-逐步回归法的泵站位移统计模型拟合效果较好,拟合精度较高,为泵站监测资料分析提供了新思路。

Finite element method and stepwise regression method-based method for constructingdisplacement statistical model of pumping station

As most of the foundations for pumping station are not consisted of rock,the more accurate statistical model for displacement of pumping station cannot be obtained through conventional mechanical methods,and then the calculation is more difficult. Therefore,a finite element—stepwise regression analysis-based method for constructing displacement statistical model of pumping station is proposed herein in accordance with the construction concept for constructing dam deformation statistical model. Based on the stepwise regression,this method determines all the water pressure-related factors in the statistical model through the finite element method at first,and then determines all the factors in the temperature component and the time-effect component according to the relevant monitoring data,thus obtains more accurate statistical model of pumping structure displacement. Through the verification made on a pumping station of the East Ｒoute of the South-to-North Water Diversion Project,it is indicated that the measured horizontal and vertical displacements are basically coincided with the results got from the 3 － D finite element— stepwise regression analysis,which is further demonstrated that it is feasible to use the statistical model for the analysis on the displacement of pumping station. The finite element—stepwise regression method-based statistical model for pumping station displacement has better fitting effect and higher fitting precision,and then provides a new idea for the analysis on the monitoring data of pumping station.