改进马尔可夫链降雨量预测模型的应用  

Precipitation Predicting Model Based on Improved Markov Chain

作　　者：仲远见[1] 李靖[1] 王龙[1] 

机构地区：[1]云南农业大学水资源与节水灌溉重点实验室,云南昆明650201

出　　处：《济南大学学报（自然科学版）》2009年第4期402-405,共4页Journal of University of Jinan(Science and Technology)

基　　金：联合国人居署亚洲城市水资源计划(UN-WCDM07A05);云南省高校学术带头人基金;云南省干热地区水资源特性研究课题(YSZY-2003-02);云南省应用基础研究面上项目(2007D210M)

摘　　要：根据马尔可夫链预测方法,汇总分析1956-1999年昆明市东川区的年降雨量资料,采用均值-标准差分级法,对各年的降雨量进行状态分级,分为洪涝、偏涝、正常、偏旱、干旱5个状态。应用马尔可夫链及转移概率矩阵对昆明市东川区2000年的降雨量进行分析预测,并根据模糊集理论中的级别特征值预测降雨量。结果表明：采用该方法预测的结果与实际情况相吻合。Basing on the precipitation data from 1956 to 1999, and taking the means and standard deviation of information series as the classification standard of precipitation ,the past 44 years are classified into five classes according to the precipitation data,i, e. ,water-logging year, weak water- logging year, normal year, weak drought year, and drought year. The precipitation of Dongchuan district of Kunming in 2000 is predicted and analyzed using the weighted Markov chain and transition probability. The level characteristics value of fuzzy sets is used to predict the concrete value of precipitation. It is concluded that the result of prediction by the present method agrees with the reality.

关 键 词：马尔可夫链 云南干热河谷区 年降雨量 模糊集理论 平稳分布 

分 类 号：TV213.9[水利工程—水文学及水资源]