渗压监测统计模型欠拟合问题及因子的扩充  

Insufficiency of Statistical Regression Model for Seepage Pressure Monitoring and Expansion of Factors

作　　者：李雅静[1] 张志诚[1] 陈广洲[2] 

机构地区：[1]河海大学水资源高效利用与工程安全国家工程研究中心,南京210098 [2]广东省水利电力勘测设计研究院,广州510635

出　　处：《长江科学院院报》2009年第6期13-15,共3页Journal of Changjiang River Scientific Research Institute

摘　　要：在使用统计回归模型来建立安全监控模型时,选用有限的因子进行拟合,往往会出现由于因子选择不确切或者不全面导致模型出现欠拟合问题,影响其拟合和预报精度,因此有必要对模型进行补充和完善。结合渗压监测的原型观测资料,用因子扩充和组合进行建模,计算结果表明这样建立的模型回归精度更高。Only finite factors are selected when using a statistic regress model establishes a .safety monitoring model. Sometimes this will lead to ambiguous or missing phenomenon of factors and the defect of statistical regression model in precision. Therefore, it is necessary to reinforce and consummate the regression model. This text takes the statistical regression model after factor expansion as reference to establish the safety monitoring model with the observed data in prototype. The results indicate this method will result in higher regression precision.

关 键 词：统计回归模型 安全监控 原型观测资料 回归精度 

分 类 号：TV698.12[水利工程—水利水电工程]