-
摘要: 针对旅客个性化出行需求和机场快速疏解要求，在对陆侧交通各出行方式固定分配的基础上，提出了一种基于二次诱导的群体空港旅客出行推荐方法，为定制化旅客服务提供算法支撑；基于旅客原始数据，结合粗糙集理论进行特征属性的知识约简，提高算法性能；运用改进贝叶斯分类算法进行基于旅客独立特征概率计算的出行方式推荐度量化，生成基于一次诱导的旅客出行推荐序列；面向机场陆侧各出行方式运力固定分配的约束，将旅客出行推荐序列输入基于改进非支配排序遗传算法(NSGA-Ⅱ)的旅客二次诱导出行推荐模型中，进行运力与旅客流的深度匹配，对旅客出行推荐结果进行再次优化；基于普适性原则，使用小规模(100人次)和大规模(1 000人次) 旅客样本进行模型验证。分析结果表明：在不同规模旅客流输入情况下均能得到良好结果，小规模样本下旅客出行方式推荐正确率为77.41%，大规模样本下旅客出行方式推荐正确率为79.62%；经过二次诱导后，旅客流出行推荐分布与运力间的匹配度相比真实出行以及一次诱导分布皆有巨大提升。在旅客流与运力高度匹配的基础上实现了旅客出行偏好需求，算法性能良好，为改善枢纽机场客流疏解能力提供了一种切实可行的方法。
-
关键词:
- 旅客出行推荐 /
- 改进非支配排序遗传算法 /
- 二次诱导 /
- 属性约简 /
- 旅客流与运力匹配
Abstract: In view of the personalized travel needs of passengers and the requirements of rapid airport evacuation, a travel recommendation method for group airport passengers based on the secondary induction was proposed on the basis of the fixed allocation of each travel mode of landside transportation, so as to provide algorithmic support for customized passenger services. Based on the original passenger data, and combined with the rough set theory, the knowledge reduction of feature attributes was carried out to improve the performance of the algorithm. The improved Bayesian classification algorithm was used to quantify the travel mode recommendation degree based on the calculation of the independent feature probability of passengers, and the passenger travel recommendation sequence based on the primary induction was generated. In view of the constraint of fixed capacity allocation of each travel mode on the landside of the airport, the passenger travel recommendation sequence was input into the secondary-induced travel recommendation model of passengers based on the improved non-dominated sorting genetic algorithm (NSGA-Ⅱ) to deeply match the transport capacity and passenger flow, and the passenger travel recommendation results were optimized again. Based on the principle of universality, the small-scale (100 people) and large-scale (1 000 people) passenger samples were used for model validation. Analysis results show that good results can be obtained under the inputs of passenger flows with different scales. The correct rate of passenger travel mode recommendation in the small-scale sample is 77.41%. Under the large-scale sample, the correct rate of passenger travel mode recommendation is 79.62%. After the secondary induction, the matching degree between the recommended travel distribution of passenger flow and the transport capacity greatly improves compared with the real travel and the primary induction distribution. On the basis of high matching between the passenger flow and the transport capacity, the passenger travel preference needs are realized. The algorithm has good performance and provides a practical method to improve the passenger flow evacuation of hub airports. -
图 4 不同约简流程的一次诱导正确率对比
Figure 4. Comparison of primary induction accuracies of different reduction processes
图 7 最优出行组合旅客流推荐分布
Figure 7. Recommended distributions of optimal travel combination passenger flows
图 8 不同诱导阶段旅客出行方式分布综合对比
Figure 8. Comprehensive comparison of distributions of passenger travel modes at different induction stages
表 1 属性约简算法结果
Table 1. Result of attribute reduction algorithm
约简算法 付款人 准时重要性 理想出发时间 理想到达时间 旅行人数 年龄 收入 性别 职业 受教育程度 出发时间 到达时间 飞行时间 旅行时间 机票价格 无约简 √ √ √ √ √ √ √ √ √ √ √ √ √ √ √ 基于依赖度的属性约简 · · — — — — √ √ — √ √ · — — √ 基于正域的属性约简 √ · — √ — — √ √ √ √ √ √ — — √ 
下载: 导出CSV
表 2 特征属性权值
Table 2. Feature attributes weights
特征属性 付款人 准时重要性 收入 性别 受教育程度 出发时间 到达时间 机票价格 权值 1.000 1.000 0.548 0.216 0.436 0.371 1.000 0.374
下载: 导出CSV
表 3 旅客出行推荐序列分布
Table 3. Distributions of passenger travel recommendation sequence
样本 推荐梯度次数 出租车/网约车 机场巴士 轨道交通 私家车 小规模
旅客样本首位 12 10 56 15 次位 28 36 16 13 三位 13 29 10 41 末位 40 18 11 24 大规模
旅客样本首位 89 65 628 149 次位 320 329 108 174 三位 88 267 84 492 末位 395 240 167 129
下载: 导出CSV
表 4 二次诱导算法参数设置
Table 4. Parameter setting of secondary induction algorithm
参数 取值 最佳运力分配比例{t1, t2, t3, t4} {0.150, 0.519, 0.161, 0.170} 进化代数N 小规模旅客样本 20、50、80 大规模旅客样本 50、100、150 基因数目na 小规模旅客样本 100 大规模旅客样本 1 000 种群数目n 5 000 目标函数数目v 2 交叉参数p1, off 1.00 交叉参数p2, off 0.60 变异参数p1, var 0.50 变异参数p2, var 0.15
下载: 导出CSV
表 5 不同诱导阶段旅客出行分布与运力适应度对比
Table 5. Comparison of adaptabilities between passenger travel distribution and transport capacity at different induction stages
样本 真实出行分布与运力适应度 一次诱导分布与运力适应度 二次诱导分布与运力适应度 小规模旅客样本 -14.89 -10.01 -4.39 大规模旅客样本 -181.83 -108.77 -37.19
下载: 导出CSV
-
[1] HE Yu-gang. Civil aviation transportation industry and economic growth: an investigation from China[J]. Journal of Tourism and Industry Research, 2018, 38(4): 5-14. [2] CHOO S, YOU S, LEE H. Exploring characteristics of airport access mode choice: a case study of Korea[J]. Transportation Planning and Technology, 2013, 36(4): 335-351. doi: 10.1080/03081060.2013.798484 [3] TABARES D A. An airport operations proposal for a pandemic-free air travel[J]. Journal of Air Transport Management, 2021, 90: 101943. doi: 10.1016/j.jairtraman.2020.101943 [4] 赵嶷飞, 肖瞳瞳, 万俊强. 中国民航空中交通管制体制演化[J]. 交通运输工程学报, 2020, 20(2): 100-120. doi: 10.19818/j.cnki.1671-1637.2020.02.009ZHAO Yi-fei, XIAO Tong-tong, WAN Jun-qiang. Evolution of air traffic control system of Chinese Civil Aviation[J]. Journal of Traffic and Transportation Engineering, 2020, 20(2): 100-120. (in Chinese) doi: 10.19818/j.cnki.1671-1637.2020.02.009 [5] 刘晓明, 夏洪山. 基于逆向重力模型的枢纽机场旅客吞吐量估计方法[J]. 交通运输工程学报, 2008, 8(2): 85-89. doi: 10.3321/j.issn:1671-1637.2008.02.018LIU Xiao-ming, XIA Hong-shan. Estimating methods of passenger throughput for hub airport based on reverse gravity model[J]. Journal of Traffic and Transportation Engineering, 2008, 8(2): 85-89. (in Chinese) doi: 10.3321/j.issn:1671-1637.2008.02.018 [6] 李国栋, 胡雅茹. 中国民航业发展与人口布局均衡程度实证研究[J]. 综合运输, 2018, 40(1): 11-16, 67.LI Guo-dong, HU Ya-ru. Empirical study on the balance degree between China's civil aviation industry development and population distribution[J]. China Transportation Review, 2018, 40(1): 11-16, 67. (in Chinese) [7] LIU Yu, WANG Wei-jie, HUANG Hong-zhong, et al. A new simulation model for assessing aircraft emergency evacuation considering passenger physical characteristics[J]. Reliability Engineering and System Safety, 2014(121): 187-197. [8] FAN Ai-hua, CHEN Xu-mei, WANG You-an, et al. All-stop, skip-stop, or transfer service: an empirical study on preferences of bus passengers[J]. IET Intelligent Transport Systems, 2018, 12(10): 1255-1263. doi: 10.1049/iet-its.2018.5213 [9] 阚劲军, 周中雨. 四型机场与四强空管有机统一浅谈[J]. 民航学报, 2021, 5(4): 29-31.GAN Jin-jun, ZHOU Zhong-yu. On organic unity of airport with four features and air traffic control with four strengths[J]. Journal of Civil Aviation, 2021, 5(4): 29-31. (in Chinese) [10] WU Xiao-kun, ZHAO Tian-fang, CHEN Wei-neng, et al. Toward predicting active participants in tweet streams: a case study on two civil rights events[J]. IEEE Transactions on Knowledge and Data Engineering, 2020, DOI: 10.1109/TKDE.2020.3017635. [11] 刘晓飞, 朱斐, 伏玉琛, 等. 基于用户偏好特征挖掘的个性化推荐算法[J]. 计算机科学, 2020, 47(4): 50-53.LIU Xiao-fei, ZHU Fei, FU Yu-chen, et al. Personalized recommendation algorithm based on user preference feature mining[J]. Computer Science, 2020, 47(4): 50-53. (in Chinese) [12] WU Cheng-lang, CHEN Yi-meng. Effects of passenger characteristics and terminal layout on airport retail revenue: an agent-based simulation approach[J]. Transportation Planning and Technology, 2019, 42(2): 167-186. doi: 10.1080/03081060.2019.1565163 [13] WANG Liu, ZHANG Yong, ZHAO Xia, et al. Irregular travel groups detection based on cascade clustering in urban subway[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 21(5): 2216-2225. doi: 10.1109/TITS.2019.2933497 [14] 原福永, 冯静, 符茜茜. 微博用户的影响力指数模型[J]. 现代图书情报技术, 2012(6): 60-64.YUAN Fu-yong, FENG Jing, FU Qian-qian. Influence index model of micro-blog user[J]. New Technology of Library and Information Service, 2012(6): 60-64. (in Chinese) [15] QIU Ying-hui, HE Zhen-huan. A research on passenger's travel choice behavior in high-speed railway based on passenger survey questionnaire[C]//IEEE. 2nd International Conference on Applied Mathematics, Modeling and Simulation (AMMS). New York: IEEE, 2018: 277-283. [16] GOSLING G D. Use of air passenger survey data in forecasting air travel demand[J]. Transportation Research Record, 2018(2449): 79-87. [17] ARAGHI Y, KROESEN M, MOLIN E, et al. Revealing heterogeneity in air travelers' responses to passenger-oriented environmental policies: a discrete-choice latent class model[J]. International Journal of Sustainable Transportation, 2016, 10(9): 765-772. doi: 10.1080/15568318.2016.1149645 [18] WEI Wei, WANG Cheng. Design of air passenger travel choice intention prediction system based on deep learning[J]. Scientific Programming, 2022, 2022: 7340552. [19] 钟翔, 韩旭, 朱彩云, 等. 基于二分K均值算法的机场旅客群体划分模型[J]. 中国民航大学学报, 2018, 36(3): 37-40.ZHONG Xiang, HAN Xu, ZHU Cai-yun, et al. Airport passenger grouping model based on dichotomic K-means algorithm[J]. Journal of Civil Aviation University of China, 2018, 36(3): 37-40. (in Chinese) [20] 吴麟麟, 卢海琴, 汪洋, 等. 引入忠诚度变量的城际出行方式动态选择行为研究[J]. 公路交通科技, 2014, 31(11): 123-129.WU Qi-lin, LU Hai-qin, WANG Yang, et al. Research on intercity travel mode dynamic choice behavior with introduced loyalty variable[J]. Journal of Highway and Transportation Research and Development, 2014, 31(11): 123-129. (in Chinese) [21] 尹溪琛. 基于旅客画像和出行链模型的交通出行推荐方法及应用[D]. 北京: 北京交通大学, 2021.YIN Xi-chen. Travel recommendation method and application based on passenger profile and trip chain model[D]. Beijing: Beijing Jiaotong University, 2021. (in Chinese) [22] WU Xiang-juan, WANG Yu-ping, WANG Zi-qing. A centerline symmetry and double-line transformation based algorithm for large-scale multi-objective optimization[J]. Connection Science, 2022, 34(1): 1454-1481. doi: 10.1080/09540091.2022.2075828 [23] MARTINEZ-ALVAREZ A, CRESPO-CANO R, DIAZ-TAHOCES A, et al. Automatic tuning of a retina model for a cortical visual neuroprosthesis using a multi-objective optimization genetic algorithm[J]. International Journal of Neural Systems, 2016, 26(7): 1650021. doi: 10.1142/S0129065716500210 [24] OLIVETO P S, SUDHOLT D, ZARGES C. On the benefits and risks of using fitness sharing for multimodal optimisation[J]. Theoretical Computer Science, 2018, 773(2019): 53-70. [25] ZHU Shu-wei, XU Li-hong, GOODMAN E D, et al. A new many-objective evolutionary algorithm based on generalized pareto dominance[J]. IEEE Transactions on Cybernetics, 2022, 52(8): 7776-7790. doi: 10.1109/TCYB.2021.3051078 [26] YUAN Yuan, XU Hua, WANG Bo, et al. A new dominance relation-based evolutionary algorithm for many-objective optimization[J]. IEEE Transactions on Evolutionary Computation, 2016, 20(1): 16-37. doi: 10.1109/TEVC.2015.2420112 [27] 刘寅斌, 余咪咪, 李洪波, 等. 多目标项目调度研究综述: 复杂系统优化的视角[J]. 系统科学学报, 2020, 28(3): 84-89, 111.LIU Yin-bin, YU Mi-mi, LI Hong-bo, et al. A literature review of multi-objective project scheduling: a perspective on complex system optimization[J]. Journal of Systems Science, 2020, 28(3): 84-89, 111. (in Chinese) [28] 张泉峰. 首都机场接续运输协调保障技术研究与实现[D]. 成都: 电子科技大学, 2015.ZHANG Quan-feng. The research and implementation of capital International Airport roadside passenger transportation techniques[D]. Chengdu: University of Electronic Science and Technology, 2015. (in Chinese) [29] 聂磊, 高艺, 佘亮. 首都机场快线客运需求预测与运营方式研究[J]. 交通运输系统工程与信息, 2009, 9(4): 151-158.NEI Lei, GAO Yi, SHE Liang. Study on the forecast of the passenger flow to Beijing Airport by airport express line and its operation capital plan[J]. Journal of Transportation Systems Engineering and Information Technology, 2009, 9(4): 151-158. (in Chinese) [30] DEB K, PRATAP A, AGARWAL S, et al. A fast and elitist multiobjective genetic algorithm: NSGA-Ⅱ[J]. IEEE Transactions on Evolutionary Computation, 2002, 6(2): 182-197. -
点击查看大图
图(8) / 表(5)
计量
- 文章访问数: 104
- HTML全文浏览量: 9
- PDF下载量: 50
- 被引次数: 0
