季节性自回归模型适用性分析  

An Analysis of the Applicability of the Seasonal Autoregressive Model

作　　者：武兰婷 何士华[1] 朱剑军[1] 裴英俊 

机构地区：[1]昆明理工大学电力工程学院,昆明650500

出　　处：《中国农村水利水电》2014年第12期17-19,23,共4页China Rural Water and Hydropower

基　　金：国家自然科学基金资助项目(51369013);昆明理工大学自然科学基金资助项目(kkz3201304016)

摘　　要：月径流序列的随机模拟通常采用季节性自回归模型[SAR(p)],该模型考虑了月径流随周期变化的特性。以仁里水文站37年的实测月径流序列为基础,计算分析其统计特性参数,根据序列中各月自回归系数的不同,对不同月取不同阶数的SAR(p)模型,并根据当地资料采用适线法估计偏态系数Cs值,建立混合SAR(p)模型。与SAR(1)模型和SAR(2)模型的模拟结果对比,混合SAR(p)模型的模拟结果更接近实测资料,偏态系数Cs值和二阶自相关系数明显优于一、二阶模型,能较好的模拟原序列的统计特性。The seasonal autoregressive model[SAR（p）]is often used for random simulation of monthly runoff.The model can keep the period change feature of monthly runoff series.The paper analyzes the statistical parameters of 37years＇ monthly runoff data of Renli Hydrometric Station,the regression coefficients are different in different months,so different months should take different order of SAR（p）model and coefficient of skewness（Cs）should be estimated according to the local data,mixed SAR（p）model is established.The mixed SAR（p）model compare with SAR（1）model and SAR（2）model,it is closer to the measured data,the value of Cs and the second order autocorrelation coefficient are better than SAR（1）model and SAR（2）model,it can better keep the statistical characteristics of the original sequence.

关 键 词：径流模拟 自回归模型 季节性自回归模型 适用性分析 

分 类 号：TV121[水利工程—水文学及水资源] P338.2[天文地球—水文科学]