基于GWO-HMM的空中交通网络流系统态势预测研究  

Research on prediction of air traffic network flow system situation based on GWO-HMM

作　　者：张兆宁[1] 杨刚 ZHANG Zhaoning;YANG Gang(College of Air Traffic Management,CAUC,Tianjin 300300,China)

机构地区：[1]中国民航大学空中交通管理学院,天津300300

出　　处：《中国民航大学学报》2024年第4期50-55,共6页Journal of Civil Aviation University of China

基　　金：国家重点研发计划项目(2020YFB1600103)。

摘　　要：针对空中交通流量管理部门如何更高效地实施流量管理的问题,本文将态势感知理论应用于空中交通网络流系统(ATNFS,air traffic network flow system),建立空中交通网络流系统的运行态势预测模型。首先,给出了空中交通网络流系统的态势感知过程,从节点和航线的角度筛选出航线饱和度、不正常航班率、节点饱和度、节点延误架次比、节点航班取消率5个态势要素,使用态势值作为态势理解的指标;其次,分析隐马尔可夫模型(HMM,hidden Markov model)的优势与不足,建立了基于灰狼优化(GWO,grey wolf optimization)算法和改进隐马尔可夫模型的态势预测模型;最后,使用某空中交通网络流系统的实际运行数据进行算例验证。结果表明,改进后的预测模型相较于原本的隐马尔可夫预测模型精度更高,预测结果更准确。To address the problem of how air traffic flow management departments can implement flow management more efficiently,the situational awareness theory was applied to the air traffic network flow system(ATNFS)in this paper,and the operational situation prediction model of the air traffic network flow system was established.Firstly,the situational awareness process of air traffic network flow system was provided,and five situation elements including route saturation,irregular flight rate,node saturation,node delayed sortie ratio and node flight cancellation rate,were selected from the perspective of nodes and routes,and the situation values were used as indicators of situation understanding.Secondly,the advantages and disadvantages of hidden Markov model(HMM)were analyzed,and a situation prediction model based on grey wolf optimization(GWO)algorithm and improved HMM was established.Finally,the actual operation data of an air traffic network flow system were used to verify the algorithm.The results showed that the improved prediction model had higher accuracy and more accurate prediction results compared with the original HMM.

关 键 词：空中交通流量管理 空中交通网络流系统 隐马尔可夫模型(HMM) 灰狼优化(GWO)算法 态势感知 态势预测 

分 类 号：V355.2[航空宇航科学与技术—人机与环境工程]