非参数解集模型再探  

Further Study on Non-parametric Disaggregation Model

作　　者：王文圣[1,2] 向红莲[2] 

机构地区：[1]中国气象局成都高原气象研究所,成都610072 [2]四川大学水利水电学院,成都610065

出　　处：《成都工业学院学报》2015年第4期47-49,共3页Journal of Chengdu Technological University

基　　金：国家自科基金项目"水文水资源集对分析方法的科学基础及应用研究"(51179110)

摘　　要：参数解集模型具有固定的相依结构和边缘概率分布,非参数解集模型克服了参数解集模型的缺点,但非参数解集模型有自身的不足:用于时间量随机解集时,不能保持首尾自相关结构一致性;用于空间量分解时,不能保持空间上的量滞时为1的自、互相关特性。基于改变条件概率密度函数对非参数解集模型进行了改进,提出了改进的非参数解集模型。将建议模型应用于金沙江流域屏山站月径流随机生成,结果表明改进模型是合适的。Disaggregation model can preserve variance, covariance and other statistical properties between lover-level variables as well as those between lower-level and higher-level variables in time or space scale.Traditional parametric disaggregation model has some drawbacks due to some prior assumptions as to form of dependence and the form of probability density function.To avoid the assumptions and overcome some drawbacks of the parametric disaggregation model, the nonparametric disaggregation model is presented.The nonparametric disaggregation model has shortcomings too.It cannot preserve back and forth auto-correlation structure of variables in time, and the auto-correlation and cross-correlation structure of variables in space.In this paper a modified nonparametric disaggregation model is proposed and applied to monthly stream simulation at Ping Shan station in Jing Sha River.The results show that the proposed model is functional.

关 键 词：解集模型 改进的非参数解集模型 模型检验 

分 类 号：P458[天文地球—大气科学及气象学]