摘要
航班延误是目前航空运输业发展所面临的一大难题。从航班延误链式波及反应的角度出发,将贝叶斯推理应用于航班过站时间的分析,分别建立实际航班数据的单机场和多机场过站时间贝叶斯网络模型。模型清晰呈现了机型类型、前航延误时间等因素对机场过站时间的影响以及首发延误等级、经停机场过站时间调整量等因素对末班延误的影响。通过从多角度对模型进行分析,结果表明发生前航延误时调整航班在机场的过站时间可以有效减少延误向下游机场的波及。
Flight delay is a major challenge in the industry of civil aviation. Spreading from the chain reaction of flight delay propagation, this paper applied Bayesian inference in analysis of aircraft turnaround time, and established two Bayesian Network models according to actual flight data:single airport model and multi- airport model. The models clearly presented the impact of aircraft type and upstream delay on turnaround time as well as the impact of starting delay and turnaround time adjustment on last flight delay. After analyzing the two Bayesian Network models from different aspects, the results show that downstream flight delay can be effectively reduced by adjusting aircraft turnaround time when the upstream flight is known delayed.
出处
《航空计算技术》
2010年第5期5-9,14,共6页
Aeronautical Computing Technique
基金
国家863重点项目(2006AA12A106)
国家自然科学基金资助项目(60572167
60879015)