基于移动平均-马尔可夫链模型的上海市机场旅客吞吐量预测研究  

Research on Shanghai airport passenger throughput forecast by the Moving Average-Markov chain model

作　　者：申瑞娜[1] 樊重俊[1] 

机构地区：[1]上海理工大学管理学院,上海200093

出　　处：《科技与管理》2014年第5期51-54,59,共5页Science-Technology and Management

基　　金：上海市教育委员会科研创新重点项目(14ZZ131);上海机场(集团)有限公司科研项目

摘　　要：为了更多的挖掘数据的信息,本文将移动平均与马尔可夫链模型结合,对马尔可夫链模型进行改进,以拟合结果为动态基准线来构建动态的马尔可夫链,并进行预测及平稳分布研究。利用上海市机场2008年1月至2013年10月的各月数据进行模型拟合,将拟合结果作为动态基准线,进行系统状态划分,确定状态转移概率矩阵。预测未来4个月的中值和区间,并根据马尔可夫链的遍历性和稳定性求出状态的平稳分布。实证表明,该方法的预测精度更高,并且具有较强的适用性和泛化性,具有较强的实用价值。For more information on data mining, this paper is to study the moving average in conjunction with Mark- ov chain model, improving Markov chain model, and treating the fitting results as a dynamic baseline to improve Markov chain, researching and forecast stationary distribution. This paper uses the data of Shanghai airport from January 2008 to October 2013 every month to fit the model, treating the fitting results as a dynamic baseline, divid- ing the system state, and defining the state transition probability matrix. It also Predicted the values and the interval for the next four months, and calculated the stationary distribution in accordance with ergodicity and stability of Markov chain. Empirical evidence shows that this method is more accurate prediction, holding strong applicability and generalization and a strong practical value.

关 键 词：旅客吞吐量 马尔可夫链模型 移动平均-马尔可夫链预测模型 

分 类 号：O211.62[理学—概率论与数理统计] U8[理学—数学]