一种适用于大坝安全监测资料分析建模的偏最小二乘回归法及其计算机实现  

A Partial Least Squares Regression Method Applied to Dam Safety Monitoring Modeling and Its Realization on Computer

作　　者：邱小弟[1] 潘琳[2] 

机构地区：[1]云南华能澜沧江水电有限公司,云南省昆明市650000 [2]国电自动化研究院/南瑞集团公司,江苏省南京市210003

出　　处：《水电自动化与大坝监测》2005年第6期58-61,共4页HYDROPOWER AUTOMATION AND DAM MONITORING

摘　　要：简要叙述了偏最小二乘回归法的基本思路和计算方法，阐明了将该方法应用于大坝安全监 测资料的分析建模，能够突破传统回归方法的局限性，获得完整、稳健的监测量模型。介绍了一套 应用程序，并通过计算实例的对比，说明了偏最小二乘回归法较传统的多元线性回归方法的优越 性，展示了其良好的应用前景。The basic idea and calculation process of the partial least squares regression method are described briefly. It is pointed out that the application of the method to dam safety monitoring modeling can overcome the limitation of traditional regression methods, and generate holistic and steady models. An application program is introduced. The comparison between some examples shows the advantages and nice future of the method.

关 键 词：偏最小二乘回归法 多重相关性 主成分分析 典型相关分析 

分 类 号：TV698.1[水利工程—水利水电工程]