基于混合正态分布的降水量分布特征估计  

Estimation of Precipitation Distribution Characteristics Based on Mixed Normal Distribution

作　　者：李娇华 LI Jiaohua(School of Mathematis and Statistics,Guangxi Normal University,Guilin 541004,China)

机构地区：[1]广西师范大学数学与统计学院,广西桂林541004

出　　处：《广西民族大学学报（自然科学版）》2022年第4期77-81,共5页Journal of Guangxi Minzu University ：Natural Science Edition

基　　金：国家自然科学基金项目(11901124);广西科技计划项目(桂科AD21220114);广西高校中青年教师基础能力提升项目(2021KY1798)。

摘　　要：降水量分布特性研究一直是水能资源研究评估的重点。文章将借助混合正态分布对年降水量的分布特征进行探讨,并利用EM算法对混合正态模型中的参数进行估计;通过模拟和实证分析对混合正态分布模型和正态分布模型的参数估计结果进行比较,结果表明混合正态分布更适合应用于水能资源分布特性的研究以及区域水能资源的评估。The study of precipitation distribution characteristics has always been the focus of hydropower resources research and evaluation.In this paper,the distribution characteristics of annual precipitation are discussed with the help of mixed normal distribution,and the EM algorithm is used to estimate the parameters in the mixed normal model;Through simulation and empirical analysis,the parameter estimation results of the mixed normal distribution model and the normal distribution model are compared.The results show that the mixed normal distribution is more suitable for the study of the distribution characteristics of hydropower resources and the evaluation of regional hydropower resources.

关 键 词：降水量 混合正态分布 参数估计 EM算法 

分 类 号：O212[理学—概率论与数理统计]