逐步回归法在渗流观测资料分析中的应用

逐步回归法能较好处理初选因子之间多重共线性问题，有效解决随机变量之间相关关系，确定显著变量因子。本文基于逐步回归法，结合实测资料进行回归分析，分析结果表明：利用逐步回归方法建立的回归模型拟合效果较好，复相关系数高，能较好反映水库大坝不同部位渗流显著影响因子；回归模型可用于进行水位预测，为了解大坝渗流运行状态提供指导和帮助。

Application of stepwise regression method in seepage analysis

The stepwise regression can apply to deal with the problem of multiple co-linearity between primary factors,and solxe the correlation between random xariables,and determining the significant xariables. Based on the stepwise regression and combined analyzing measured data regression,the results showed that the fitting effect stepwise regression model is well,multiple correlation coefficient is large,reflecting the different significant influence factor and dexelop trend,predicting the water lexel with the regression model can proxide guidance for the understanding of the status of the dam seepage.