基于多目标优化的改航策略分析-analysis of rerouting strategy based on multi-objective optimization.docxVIP

摘要随着我国民航运输业的快速发展，空域资源日益紧张，尤其是发生恶劣天气和重大事件时， 可用空域资源将会急剧减少，继而导致航班延误、旅客滞留和航班取消，引起严重的经济损失 和较大的安全隐患。改航作为空中交通流量管理的有效手段，主要研究容量受限下的经济损失 最小化问题。目前主要是研究单目标优化的改航策略，并使用人工估计的方法评估受影响的扇 区动态容量，这种改航方法降低空域资源的使用效率的同时也缺乏科学性，因此亟需研究基于 多目标优化的改航策略。本文在前期学者研究的基础上，对基于多目标优化的改航策略展开了新的研究。首先分析 了恶劣天气和重大事件对飞行的影响，继而结合我国民航气象系统中恶劣天气的预报和发布方 式以及重大事件预报机制，分别建立了飞行受限区。然后研究了恶劣天气和重大事件对扇区动 态容量的影响并建立了扇区动态容量评估模型；在获取航班途经扇区和机场容量的基础上建立 了容量情景树，继而建立了基于多目标优化的改航模型，该模型可规划航班放行组织方案与基 于避让飞行受限区的路径选择方法。最后设计并实现了基于多目标优化的改航系统，仿真实例 验证了所提模型的有效性和实用性。 关键词：空中交通流量管理，多目标改航策略，恶劣天气，重大事件，扇区动态容量IABSTRACTWith the rapid development of the civil aviation of China, the airspace resource has become congested. The available airspace resource of civil aviation will be sharply reduced especially when severe weather and major events emerge. Then these situations may lead to a large number of flights delay, the passengers’ backup, the flights cancellation, the serious economic losses and the potential safety troubles. Rerouting strategy which is one of the effective mean of air traffic flow management has good application value in the flight planning problem in limited airspace. However the present rerouting strategy is usually based on single target, and the affected sector capacity is usually evaluated by artificial method, which not only reduces the airspace resource using efficiency, but also lacks scientific evaluation. So the multi-objective optimization rerouting strategy has become the key to solve the problem.This paper launched a new study on multi-objective optimization rerouting strategy based on former research. Firstly, the paper analyzed the effect of severe weather and major events on the scheduled flights, and then established flight forbidden areas respectively based on the severe weather forecast and major events issue mechanism. Then the paper analyzed the influence of the severe weather and major events on related sector dynamic capacity, and established a sector dynamic capacity evaluation model. Secondly, considering the secto