基于马尔可夫链的自然风风向建模  

Wind direction modeling using Markov chain

作　　者：季亮[1] 谭洪卫[1] 王亮[1] 

机构地区：[1]同济大学机械工程学院,上海200092

出　　处：《中南大学学报（自然科学版）》2012年第8期3274-3279,共6页Journal of Central South University:Science and Technology

基　　金：国家自然科学基金资助项目(50778128);高等学校博士学科点基金资助项目(20090072110016)

摘　　要：根据马尔可夫链基本理论,基于自主构建的小型气象站实测得到的高采集频度基础数据,以自然风风向为研究对象,对自然风风向建模并基于该模型生成随机时间序列,构建上海地区的自然风风向变化的转移概率概率矩阵模型。通过统计参数的分析证明:该方法对分析随机变化的风向特征具有通用性。最后,基于实测数据和该方法分析出的上海地区风向变化转移概率矩阵。Based on the fundamental theory of Markov chain, an approach for modeling wind direction was proposed. The long-term data of wind from on-site measurement were used as source data to generate a 36-status transition matrix, which was a stochastic model of wind direction. Then, artificial time series of wind direction was predicted using this approach and the 36-status transition matrix. This approach was proved to be universal for wind direction modeling and predicting by statistical analysis. Furthermore, a 36-status transition matrix of urban area of Shanghai city was generated based on this method and the measured data.

关 键 词：风向模型 马尔可夫链 预测 时间序列 

分 类 号：P425.1[天文地球—大气科学及气象学]