马斯京根汇流参数抗差估计研究  

Research on Parameter Robust Estimation for Muskingum Model

作　　者：沈丹丹[1] 包为民[1] 刘可新[1] 龚婷婷[1] 张乾[1] 陈伟东[1] 

机构地区：[1]河海大学水文水资源学院,南京210098

出　　处：《中国农村水利水电》2016年第7期72-74,78,共4页China Rural Water and Hydropower

基　　金：国家自然科学基金面上基金(51279057);水利部公益性行业科研专项项目(201501059)

摘　　要：当河段实测流量资料存在大的、不服从正态分布的粗差或极值误差时,传统的估计方法(如最小二乘法)将不再适用。将附有条件的抗差最小二乘法运用到闽江流域沙县河段理想模型汇流参数估计中,利用两种常见的误差生成模式,分析比较了最小二乘法和抗差最小二乘法的在这两种误差影响下参数估值的效果和稳定性。结果表明,抗差最小二乘法能有效抵御粗差和极值误差的影响,从而取得较为精确且稳定性好的参数估值。When the measured reach flood is seriously influenced by the coarse errors or extreme errors, which have normal distribu- tion characteristics, traditional parameter estimation methods are not applicable. In this paper, the robust estimation with conditions is applied to estimating the Muskingum parameters of the ideal model of Shaxian reach in Minjiang River. Two common error genera- tion models are used to analyze and compare the effectiveness and stability of the least squares method and the robust least squares es timation method. The results show that the robust least squares estimation method can resist the influence of coarse errors effectively and obtain more precise and stable parameter estimates.

关 键 词：汇流参数 异常误差 最小二乘法 抗差估计 

分 类 号：TV122[水利工程—水文学及水资源]