基于加权马尔科夫的航班延误预测研究

Research on Flight Delays Prediction Based on the Weighted Markov

航班延误问题是大型枢纽机场运行管理的一个难题,针对这个问题建立基于加权马尔科夫链的定性定量预测模型,在传统马尔科夫的基础上引入加权的概念,并结合定性与定量方法进行综合预测。定性预测航班延误的升、降趋势;以延误率为依据,将延误状态分为4个级别,定量预测下一时间段的延误状态。两种方法相互检验、相互补充。结合某机场实例进行验证,实验结果表明:模型计算方便,能有效预测航班延误情况,为机场管理提供决策依据。

Flight delay has been a severe problem of operation and management at large aeronautic hub.In order to solve this problem,a new model is established based on qualitative and quantitative weighted Markov chain.Introduce weighted concept into the traditional Markov for comprehensive prediction combined with qualitative and quantitative methods.Qualitative forecast focuses on the rising and falling trend of the flight delays.While based on flight delay rate,the delay state is divided into four levels to achieve quantitative forecast of the next time＇s delay state.Qualitative and quantitative forecast could get fine mutual verification and supplement.The result testing with an airport shows that this method is convenient for calculation,and also can predict the flight delays effectively.It can provide decision basis for the airport authorities.