基于改进粒子群算法的离场航班排序  

Departure Flight Sequencing Method Based on Improved PSO Algorithm

作　　者：任易航 康瑞[1] REN Yi-hang;KANG Rui(Civil Aviation Flight University of China,Guanghan 618000,China)

机构地区：[1]中国民用航空飞行学院,四川广汉618000

出　　处：《航空计算技术》2023年第2期30-34,共5页Aeronautical Computing Technique

基　　金：四川省科技厅重点研发项目资助(2021YFG0171);中央高校基本科研业务费基金项目资助(ZJ2021-05);中国民用航空飞行学院大学生创新创业项目资助(202210624003)。

摘　　要：航班离场过程中,以时间或经济损失最小的单目标排序会导致延误时间分配不均和多目标排序在求解时存在算法收敛速度慢、易于陷入局部最优的缺陷,导致计算效率低。为解决这一问题,基于航班优先级、尾流间隔、航班延误时间、航班延误标准差构造航班离场排序模型,对粒子群算法的惯性权重和学习因子采用动态调整的改进策略。以厦门高崎机场非拥挤和拥挤场景下的共4 h离场航班进行优化排序验证,结果表明:与先到先服务(FCFS)方法、惯性权重线性递减粒子群(LDWPSO)算法相比,文中方法非拥挤场景下延误总时间减少了72%,26%,延误标准差减少了27%,28%;拥挤场景下,较FCFS延误总时间减少69%,延误标准差减少68%,与LDWPSO算法相比,优化效果上无明显差异,但在解空间的迭代收敛速度更快,达到最优罚值的迭代速度提升了55.6%。In order to solve the flight departure problem that single objective sorting with the least time or economic loss will lead to uneven delay time distribution and multi objective sorting defects that resulting in low computational efficiency,a departure sorting model is constructed based on flight priority,wake separation,flight delay time and flight delay standard deviation.An improved strategy of dynamic adjustment is adopted for the inertial weights and learning factor of PSO algorithm,combined with the 4 hours departure flight of Xiamen airport.Simulation results show that:compared with the FCFS method and LDWPSO algorithm,the total delay time under non congested scenarios is reduced by 72%and 26%,and the standard deviation of delay is reduced by 27%and 28%;under congested scenarios,the total delay time of this method is reduced by 69%compared with FCFS method.Compared with the LDWPSO algorithm,the iterative speed of the proposed algorithm in solution space is increased by 55.6%.

关 键 词：空中交通管理 航班延误公平性 惯性权重 粒子群算法 

分 类 号：O242[理学—计算数学] V352[理学—数学]