基于混合Transformer的对外客运枢纽抵站客流预测模型

doi:10.13229/j.cnki.jdxbgxb.20231091

Abstract

Abstract:

The impact of various external factors such as holidays， weather conditions， and epidemic prevention policies was considered in this paper， based on the passenger arrivals date at Beijing Capital International Airport between March 2021 and May 2023， and seven major high-speed railway stations from December 2021 to July 2023， a hybrib Transformer model integrating convolutional neural networks， long short-term memory networks is proposed for hab-to-station passenger flow prediction， analyzing the time-varying patterns of passenger flow volume at arrival stations under the influence of multiple factors. The empirical results reveal that the hybrib Transformer model outperforms conventional machine learning and deep learning models in prediction accuracy， achieving over 80% across all stations， and nearing 90% for six of them， this evidences the model's robust applicability.

Key words: engineering of communications and transportation system, external transportation hubs, intercity mobility, passenger flow prediction, hybrid Transformer model

CLC Number:

U125

Jiang-bo YU,Jian-cheng WENG,Peng-fei LIN,Yu-xing SUN,Jiao-long CHAI. Passenger flow prediction model of external transportation hub based on hybrid Transformer[J].Journal of Jilin University(Engineering and Technology Edition), 2025, 55(7): 2251-2259.

Figures/Tables 9

Table 1

Table 2

Fig.1

Fig.2

Fig.3

Table 3

Table 4

Table 5

Fig.4

References 18

[1]	刘咏. 城市对外综合客运枢纽功能定位及相关问题研究[D]. 四川:西南交通大学交通运输学院, 2008.
	Liu Yong. Research on functional orientation and related issues of urban comprehensive passenger transport hubs[D]. Sichuan: School of Trarsportation, Southwest Jiaotong University, 2008.
[2]	Liu L, Chen R C, Zhu S. Passenger flow prediction using weather data for metro systems[C]∥Proceedings of the 2018 Conference on Technologies and Applications of Artificial Intelligence. Taichung: IEEE, 2018: 70-73.
[3]	巫威眺, 曾坤, 周伟, 等. 基于多源数据和响应面优化的公交客流预测深度学习方法[J]. 吉林大学学报:工学版, 2023, 53(7): 2001-2015.
	Wu Wei-tiao, Zeng Kun, Zhou Wei, et.al. Deep learning method for bus passenger flow prediction based on multi-source data and surrogate-based optimization[J]. Journal of Jilin University(Engineering and Technology Edition), 2023, 53(7): 2001-2015.
[4]	Guo Z, Zhao X, Chen Y, et al. Short‐term passenger flow forecast of urban rail transit based on GPR and KRR[J]. IET Intelligent Transport Systems, 2019, 13(9): 1374-1382.
[5]	Zhang J, Chen F, Cui Z, et al. Deep learning architecture for short-term passenger flow forecasting in urban rail transit[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(11): 7004-7014.
[6]	Sha S, Li J, Zhang K, et al. RNN-based subway passenger flow rolling prediction[J]. IEEE Access, 2020, 8: 15232-15240.
[7]	Jiao F, Huang L, Song R, et al. An improved STL-LSTM model for daily bus passenger flow prediction during the COVID-19 pandemic[J]. Sensors, 2021, 21(17): No.5950.
[8]	Zhang L, Yang H, Wang K, et al. The impact of COVID-19 on airline passenger travel behavior: An exploratory analysis on the chinese aviation market[J]. Journal of Air Transport Management, 2021, 95: No.102084.
[9]	Qian H, Yang Z, Weng J, et al. Short-term passenger flow prediction of metro stations around sports events based on afc data[C]∥The 20th COTA International Conference of Transportation Professionals, Online,2020:70-73.
[10]	Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
[11]	张惠臻, 高正凯, 李建强, 等. 基于循环神经网络的城市轨道交通短时客流预测[J]. 吉林大学学报:工学版, 2023, 53(2): 430-438.
	Zhang Hui-zhen, Gao Zheng-kai, Li Jian-qiang, et al. Short⁃term passenger flow forecasting of urban rail transit based on recurrent neural network[J]. Journal of Jilin University(Engineering and Technology Edition), 2023,53(2): 430-438.
[12]	Zhang X, Zhong S, Mahadevan S. Airport surface movement prediction and safety assessment with spatial–temporal graph convolutional neural network[J]. Transportation Research Part C: Emerging Technologies, 2022, 144: No.103873.
[13]	Zhang J, Che H, Chen F, et al. Short-term origin-destination demand prediction in urban rail transit systems: a channel-wise attentive split-convolutional neural network method[J]. Transportation Research Part C: Emerging Technologies, 2021, 124: No. 102928.
[14]	Cheng Z, Lu J, Zhou H, et al. Short-term traffic flow prediction: an integrated method of econometrics and hybrid deep learning[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(6): 5231-5244.
[15]	Hu S, Xiong C. High-dimensional population inflow time series forecasting via an interpretable hierarchical transformer[J]. Transportation Research Part C: Emerging Technologies, 2023, 146: No.103962.
[16]	Zhou H, Li J, Zhang S, et al. Expanding the prediction capacity in long sequence time-series forecasting[J]. Artificial Intelligence, 2023, 318: No. 103886.
[17]	Ramana K, Srivastava G, Kumar M R, et al. A vision transformer approach for traffic congestion prediction in urban areas[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(4): 3922-3934.
[18]	Fisher A, Rudin C, Dominici F. All models are wrong, but many are useful: Learning a variable's importance by studying an entire class of prediction models simultaneously[J]. Journal of Machine Learning Research, 2019, 20: 1-8.

Related Articles 15

[1]	Shu-shan CHAI,Zhi-qiang ZHOU,Hai-tao LI,Jiong-yang XU. Real-time road network traffic anomaly incident detection based on graph spatial-temporal pattern learning network [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(7): 2145-2161.
[2]	Hong-zhuan ZHAO,Ze-jian WU,Xin ZHANG,Sheng-wen SHI,Wen-yong LI,Xin ZHAN,En-yong XU,Jia-ming WANG. Curve lattice model for connected commercial vehicles based on density dispersion and information transmission delay [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(6): 2015-2029.
[3]	Yi-yong PAN,Jia-cong XU,Yi-wen YOU,Yong-jun QUAN. Multi-scale spatial heterogeneity analysis of influencing factors of ride-hailing travel demand [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(5): 1567-1575.
[4]	Kai-ming LU,Yan-yan CHEN,Yao TONG,Jian ZHANG,Yong-xing LI,Ying LUO. Data-driven prediction of departure state for tail vehicles in queues at signalized intersections [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(4): 1275-1286.
[5]	He-shan ZHANG,Meng-wei FAN,Xin TAN,Zhan-ji ZHENG,Li-ming KOU,Jin XU. Dense small object vehicle detection in UAV aerial images using improved YOLOX [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(4): 1307-1318.
[6]	Jiao-rong WU,Xu-dong LIU. Analysis of influence of built environment of spatial units of different housing types on commuting mode choice [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(2): 554-565.
[7]	Fa-cheng CHEN,Guang-quan LU,Qing-feng LIN,Hao-dong ZHANG,She-qiang MA,De-zhi LIU,Hui-jun SONG. Review of drivers' takeover behavior in conditional automated driving [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(2): 419-433.
[8]	Yong-ming HE,Jia FENG,Kun WEI,Ya-nan WAN. Analysis on influencing factors of vehicle braking sideslip in curved section of superhighway [J]. Journal of Jilin University(Engineering and Technology Edition), 2025, 55(2): 591-602.
[9]	Jie MA,Zhi-li LIU,Shu-ling WANG,Hao DONG. Passenger flow prediction at entrance and exit of rail transit stations：a case study of Beijing [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(8): 2197-2205.
[10]	Zhao-wei QU,Lin LI,Yong-heng CHEN,Chang-jian WU. Traffic characteristics and safety analysis of long interval U-turn intersections [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(8): 2206-2213.
[11]	Yong-ming HE,Cong QUAN,Kun WEI,Jia FENG,Ya-nan WAN,Shi-sheng CHEN. Perceptual fusion method of vehicle road cooperation roadside unit in superhighway [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(7): 1923-1934.
[12]	Guo-zhu CHENG,Lin SHENG,Hao-yu WANG,Tian-jun FENG. Safety evaluation method for pedestrians crossing street at signalized intersection considering secondary conflict of right-turn vehicles [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(7): 1903-1912.
[13]	Ming-ye ZHANG,Min YANG,Yu LI,Shi-yu HUANG,Qing-yun LI. Optimal electric bus scheduling with multiple vehicle types considering coordinated recharging strategy [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1293-1301.
[14]	Da-yi QU,Ke-kun ZHANG,Yuan GU,Tao WANG,Hui SONG,Shou-chen DAI. Analysis of lane⁃changing decision⁃making behavior and molecular dynamics modeling for autonomous vehicles [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(3): 700-710.
[15]	De-lin LI,Jun-xian CHEN,Yong-gang WANG,Lu WANG,Zhao-qing SHEN. Identification of driving behavior on steep sharp curves based on latent class model [J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(12): 3526-3533.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

Comments

Recommended 0

No Suggested Reading articles found!

因素	2021/3/2	2021/3/2	2021/3/2	…
时间	8：00~9：00	9：00~10：00	10：00~11：00	…
实际抵站客流/人次	584	1 032	2 239	…
预售票/张	1 775	1 980	2 759	…
星期	2	2	2	…
节假日	4	4	4	…
月份	3	3	3	…
季节	1	1	1	…
风速/（m·s^-1）	1	2	2	…
天气	雾霾	雾霾	雾霾	…
最低温度/℃	-1.2	-1.2	-1.2	…
最高温度/℃	5	5	5	…
航班准点率/%	100	100	100	…
航班数量/架次	6	13	19	…
新增病例数	0	0	0	…
进返京限制政策	1	1	1	…

疫情防控等级	政策
第一级管控	中高风险地区所在市全域人员限制进京，中高风险地区人员不得进京。全面升级北京首都国际机场、大兴国际机场防控措施。所有人员进（返）京须持48 h内核酸检测阴性证明和“北京健康宝”绿码
第二级管控	有1例及以上本土新冠病毒感染者所在地级市（直辖市、副省级城市的县、市、区）其他县人员来京的须持“北京健康宝”绿码和登机登车前48 h内核酸检测阴性证明；高风险地区人员暂不许返京
第三级管控	国内低风险地区进返京人员仅需持健康通行码绿码即可进京；国内中高风险地区及全域实行封闭管理的人员，原则上不能进京

影响因素	相对重要度	排序
时段属性	1	1
航班数	0.92	2
预售票量	0.84	3
天气状况	0.89	4
进返京限制政策	0.85	5
本土新增新冠肺炎病例数	0.73	6
准点率	0.72	7
日期属性	0.71	8
节假日	0.65	9
风速	0.59	10
月份	0.47	11
季节	0.32	12
最低气温	0.27	13
最高气温	0	14

场站	验证指标	GBDT	LSTM	Transformer	CNN-Transformer	LSTM-Transformer	CL-Transformer
北京首都国际机场	MAE	375.98	269.8	230.6	134.6	144.8	88.6
	RMSE	432.01	359.1	462.3	196.1	191.9	134.4
	Accuracy/%	71.8	80.1	72.5	87.5	86.6	91.8
北京站	MAE	723.6	875.9	676.8	538.9	746.2	440.3
	RMSE	1 824.3	4 868.3	2 470.2	2 388.9	2 486.9	2 222.3
	Accuracy/%	43.2	66.8	74.3	79.5	71.7	83.3
北京西站	MAE	1 379.5	943.2	634.3	784.3	1 020	590
	RMSE	2 657.6	4 155.6	2 360.4	2 997.5	2 954.9	1 418.3
	Accuracy/%	51.2	82.4	88.2	85.3	81	89
北京南站	MAE	1 877.6	1 336.7	1 242.7	958.6	1 515.4	697.4
	RMSE	3 728.9	4 034.2	3 375.5	3 951.3	3 881.7	3 088.5
	Accuracy/%	39.5	79.7	81.1	85.4	77.1	89.4
北京北站	MAE	139.1	202.2	62.4	43.4	106.7	29.13
	RMSE	187.8	423.2	83.7	70.7	175.4	51.8
	Accuracy/%	54.8	64.9	89.2	92.5	81	95.9
清河站	MAE	244.6	56.4	134.3	45.1	164.3	44.7
	RMSE	420	129.8	195.7	115.2	322.5	117.6
	Accuracy/%	43.1	93.4	84.3	94.8	80.9	94.8
北京朝阳站	MAE	501.5	734.8	623.6	533.9	762	492.7
	RMSE	2 108.7	4 648.3	3 811.9	3 523.7	3 675.8	3 337.1
	Accuracy/%	37.5	61.1	65.3	68.9	55.5	81.3
北京丰台站	MAE	648.1	505.2	634.37	460.9	813.5	350.6
	RMSE	1 269.8	1 308.9	1 769.9	1 500.8	1 861.3	1 289.6
	Accuracy/%	46.6	75.7	69.6	77.9	61	83

Passenger flow prediction model of external transportation hub based on hybrid Transformer

RICH HTML

PDF (PC)