Hybrid particle swarm optimization arithmetic for recovery scheduling of flight delays
-
摘要: 为了优化航班延误恢复调度, 考虑了航班延误的经济效益、社会影响和经济损失构成, 定义了航线影响因子, 构建了一种新的航班延误恢复调度模型, 将局部搜索方法引入到粒子群算法中, 提出了求解航班延误恢复调度问题的混合粒子群算法。计算结果表明: 与先来先服务调度方法相比, 混合粒子群算法可以减少航班延误损失4.2%, 与基本粒子群算法和进化策略算法相比, 混合粒子群算法平均可减少航班延误损失2.0%, 随着航班延误恢复规模的增大, 算法优势会更明显。Abstract: In order to optimize the recovery scheduling of flight delays, airline impact factors were defined, the economic benefit, social impact and loss constitution of flight delays were considered, a new recovery scheduling model of flight delays was created, a hybrid particle swarm optimization arithmetic (HPSOA) was put forward, and local search method was introduced into the arithmetic.Computation result shows that HPSOA can reduce the flight delay losses by 4.2% compared with first-come-first-serve (FCFS) strategy, and evenly reduce the flight delay losses by 2.0% compared with basic PSOA and evolutionary strategy (ES), so the advantage of HPSOA is more obvious with the increase of recovery scale in flight delays.
-
表 1 各类飞机每小时延误运营成本
Table 1. α0 of different aircrafts
机型 代表符号 最大起飞质量/t 尾流类型 α0/ (元·h-1) 重型机 H > 136 重型尾流 4 167 中型机 M 7~136 (含) 中型尾流 2 916 轻型机 L ≤7 轻型尾流 208 表 2 国外旅客延误时间价值
Table 2. Delay time values of foreign passengers
$·h-1 旅客类型 平均时间价值 最低时间价值 最高时间价值 休闲旅客 23.3 20.0 30.0 商务旅客 40.1 32.1 48.1 表 3 各机场所对应的航线影响因子
Table 3. Airline impact factors of airports
机场名称 航线影响因子 上海虹桥机场 0.103 广州白云机场 0.059 上海浦东机场 0.049 深圳宝安机场 0.040 大连周水子机场 0.038 成都双流机场 0.041 昆明巫家坝机场 0.024 其他机场 0.007 表 4 航班数据
Table 4. Scheduled flight data
航班号 机型 计划起飞时刻 目的机场 最大载客人数 1 H 8∶00∶00 上海虹桥机场 420 2 H 8∶05∶00 上海虹桥机场 440 3 M 8∶10∶00 海口美兰机场 138 4 M 8∶10∶00 西安咸阳机场 189 5 M 8∶15∶00 海口美兰机场 189 6 M 8∶15∶00 兰州中川机场 189 7 M 8∶20∶00 昆明巫家坝机场 162 8 H 8∶25∶00 西安咸阳机场 440 9 M 8∶30∶00 福州长乐机场 189 10 H 8∶30∶00 上海虹桥机场 335 38 M 10∶00∶00 上海虹桥机场 180 39 M 10∶05∶00 广州白云机场 180 40 M 10∶05∶00 杭州萧山机场 189 41 H 10∶10∶00 大连周水子机场 290 42 M 10∶15∶00 广州白云机场 180 表 5 HPSOA和FCFS策略的排序
Table 5. Sequencings of HPSOA and FCFS strategy
表 6 HPSOA优化结果
Table 6. Optimization results of HPSOA
优化次数 总经济损失/元 计算时间/s 1 1 231 251 221 2 1 235 379 222 3 1 224 727 261 4 1 231 018 196 5 1 235 816 254 6 1 236 313 133 7 1 227 700 231 8 1 236 849 252 9 1 232 421 148 10 1 236 867 251 表 7 PSOA优化结果
Table 7. Optimization results of PSOA
优化次数 总经济损失/元 计算时间/s 1 1 250 798 48 2 1 254 921 48 3 1 255 222 38 4 1 249 071 39 5 1 258 400 46 6 1 249 601 20 7 1 254 439 37 8 1 254 224 42 9 1 253 272 52 10 1 258 059 32 表 8 ES优化结果
Table 8. Optimization results of ES
优化次数 总经济损失/元 计算时间/s 1 1 254 392 61 2 1 251 608 34 3 1 256 148 76 4 1 252 655 79 5 1 259 735 43 6 1 253 079 73 7 1 251 740 82 8 1 249 876 55 9 1 253 875 68 10 1 248 396 54 -
[1] 马正平, 崔德光. 机场航班延误优化模型[J]. 清华大学学报: 自然科学版, 2004, 44 (4): 474-477.Ma Zheng-ping, Cui De-guang. Optimizing airport flight delays[J]. Journal of Tsinghua University: Science and Technology, 2004, 44 (4): 474-477. (in Chinese) [2] 徐肖豪, 姚源. 遗传算法在终端区飞机排序中的应用[J]. 交通运输工程学报, 2004, 4 (3): 121-126.Xu Xiao-hao, Yao Yuan. Application of genetic algorithm to aircraft sequencing in terminal area[J]. Journal of Traffic and Transportation Engineering, 2004, 4 (3): 121-126. (in Chinese) [3] 王翠茹, 张江维, 王玥, 等. 改进粒子群优化算法求解旅行商问题[J]. 华北电力大学学报, 2005, 32 (6): 47-51.Wang Cui-ru, Zhang Jiang-wei, Wang Yue, et al. Modified particle swarm optimization algorithm for traveling salesman problem[J]. Journal of North China Electric Power University, 2005, 32 (6): 47-51. (in Chinese) [4] 夏蔚军, 吴智铭, 张伟, 等. 微粒群优化在Job-shop调度中的应用[J]. 上海交通大学学报, 2005, 39 (3): 381-385.Xia Wei-jun, Wu Zhi-ming, Zhang Wei, et al. Application of particle swarm optimization in the job-shop scheduling problem[J]. Journal of Shanghai Jiaotong University, 2005, 39 (3): 381-385. (in Chinese) [5] 那加. 基于自适应变异的粒子群优化算法的车间作业调度优化及其软件实现[J]. 信息与控制, 2005, 34 (3): 365-368.Na Jia. Application of particle swarm optimization with adaptive mutation to job shop scheduling problem and its software implementation[J]. Information and Control, 2005, 34 (3): 365-368. (in Chinese) [6] 彭传勇, 高亮, 邵新宇, 等. 求解作业车间调度问题的广义粒子群优化算法[J]. 计算机集成制造系统, 2006, 12 (6): 911-917.Peng Chuan-yong, Gao Liang, Shao Xin-yu, et al. General particle swarm optimization algorithm for job-shop scheduling problem[J]. Computer Integrated Manufacturing Systems, 2006, 12 (6): 911-917. (in Chinese) [7] 王素欣, 高利, 崔小光, 等. 交通分配的粒子群优化算法[J]. 交通运输工程学报, 2007, 7 (5): 97-102.Wang Su-xin, Gao Li, Cui Xiao-guang, et al. Particle swarm optimization arithmetic of traffic assignment[J]. Journal of Traffic and Transportation Engineering, 2007, 7 (5): 97-102. (in Chinese) [8] 邱玉琢, 陈森发. 综合运输通道客运结构配置的随机饱和熵模型[J]. 交通运输工程学报, 2007, 7 (2): 104-108.Qiu Yu-zhuo, Chen Sen-fa. Stochastic saturation entropy model of passenger transportation structure configuration for comprehensive transportation corridor[J]. Journal of Traffic and Transportation Engineering, 2007, 7 (2): 104-108. (in Chinese) [9] 徐肖豪, 李雄. 航班地面等待模型中的延误成本分析与仿真[J]. 南京航空航天大学学报, 2006, 38 (1): 115-120.Xu Xiao-hao, Li Xiong. Cost analysis of flight delays and simulation in ground-holding model[J]. Journal of Nanjing University of Aeronautics and Astronautics, 2006, 38 (1): 115-120. (in Chinese) [10] 张静. 2004年世界前100家航空公司经营状况[J]. 中国民用航空, 2005, 6 (11): 27-31.Zhang Jing. Operating status of world top 100 airlines 2004[J]. China Civil Aviation, 2005, 6 (11): 27-31. (in Chinese)