基于军民融合的全局飞行流量协同优化方法

A global network flight flow assignment algorithm based on civil-military integration

随着飞行活动需求的持续快速增长和空域资源使用矛盾的日益凸显,全局飞行流量协同优化已成为减少飞行延误、降低飞行危险、确保空域运行安全的一个重要手段。空中交通管理作为军民融合发展的重点领域,迫切需要对军民航飞行流量实施统一、高效、兼顾各自特点的协同优化。在实际研究中,全局飞行流量协同优化问题具有大规模、多目标、难分解等特点,是一类复杂的工程优化问题。本文贯彻军民融合发展思想,设计了一种基于军民航异质化飞行活动管制要求、考虑差异化调配方法与代价、兼顾军民航管制员各自工作特点、有效解决扇区网络运行安全性和经济性问题的全局飞行流量多目标协同优化模型——CMI模型;为解决种群在进化过程中"不平衡不充分"的问题,提出了一种动态自适应多目标遗传算法(DA-MOGA),并针对性设计了基于聚集距离和种群多样性的交叉变异概率动态调整机制。利用中国扇区网络实际数据,对本文提出的模型和算法进行了验证,算法结果优于2种经典的多目标进化算法。

With the rapidly continuing growth in demand for flight activities and the increasing airspace usage conflicts,the global optimization of air traffic flow management has become an essential approach to reduce flight delays,decrease flight risk and ensure airspace operation safety. As a typical area of civil-military integration development,air traffic management needs the uniform and efficient integration optimization of the civil and military aviation flight plans. The global optimization of air traffic flow management problem is a complex real-world optimization problem due to its large-scale and multi-objective,and nonseparable characteristics. This paper presents a civil-military integration flight flow multi-objective optimization——CMI model,which considers the difference in civil and military flight plans,the efficiency and safty of sector network,and the civil and military controllers operating features. In order to resolve the unbalance and inadequacy problem lying in population evolution process,a dynamic adaptive multi-objective genetic algorithm（ DA-MOGA）,which designs the dynamic adjustment mechanism of crossover and variation based on the crowding distance and diversity,is proposed in this paper. The validation results based on the actual data from the sector networks in China show that the DA-MOGA outperforms the two well-known multi-objective evolutionary algorithms.