流量资料抗差方法  

Robust Correction Method of Flow Data

作　　者：李佳佳[1] 包为民[1,2] 刘可新[1] 杨姗姗[1] 赵丽平[1] 

机构地区：[1]河海大学水文水资源学院,江苏南京210098 [2]河海大学水文水资源与水利工程科学国家重点实验室,江苏南京210098

出　　处：《水力发电》2016年第4期27-30,共4页Water Power

基　　金：国家自然科学基金资助项目(51279057;41371048);国家自然科学基金重大项目(51190091);国家重点实验室专项基金(2009586412);中央高校基本科研业务费专项资金资助(2014B35314)

摘　　要：流量资料是洪水预报中重要的信息,它受到不服从正态分布粗差影响时,会将粗差带入预报系统,影响预报结果精度。针对这种情况,提出了流量资料的抗差方法。该方法的关键就是确定适合流量资料抗差方法的权函数和权函数变量。列出两种权函数和四种权函数变量,通过15场洪水在理想模型中的抗差效果来分析确定适合的权函数和权函数变量,并通过理想模型验证该方法在不同误差下的抗差效果。随着误差的增大,抗差效果趋于稳定,这符合抗差估计理论。Flow data is important to flood forecasting. When it is influenced by the coarse errors which are non-normal distribution, the coarse errors will be brought into forecasting system and will cause an effect on the accuracy of forecasting results. To this question, the robust correction method of flow data is proposed, in which, the key is to determine the weighting function and variable equation. Two weighting functions and four variable equations are developed and evaluated by 15 floods in ideal model to analyze the robust effects, and the range of errors in which the method is available is testified. With the increase of errors, the robust effects tend to be steady.

关 键 词：流量资料 抗差方法 权函数 权函数变量 理想模型 

分 类 号：P33-3[天文地球—水文科学]