With the rapid development of the railway industry and the continuous increase of passenger transport tasks, railway passenger stations are facing increasingly severe passenger flow safety issues. To realize real-time monitoring of passenger flow dynamics and finely analyze the multi-granularity characteristics of passenger flow, a Multi-granularity Yardstick for Dynamic Crowds (MYDC) model for railway passenger stations based on video analysis technology is proposed. Firstly, a passenger flow dataset for railway passenger stations is constructed. Secondly, a fine-grained feature perception network for passenger flow is designed based on YOLO and Discriminative Correlation Filter (DCF) tracking algorithm, and the adaptive crowd localization Transformer (CLTR) model for railway passenger stations is improved to capture the coarse-grained features of the overall passenger flow distribution. Finally, based on the physical attributes of passenger flow as well as its micro and macro characteristics, a Multi-Attention Spatio-Temporal Graph Convolutional Network (MASTGCN) is constructed to mine the spatio-temporal dynamic trends of passenger flow and assess the safety risk level of passenger flow in the station. The results show that the cumulative error of fine-grained feature extraction is 6.9%, the recognition accuracy of coarse-grained features is 89.1%, and the recall rate of the passenger flow safety assessment model is 87.5%. The proposed model can provide accurate data support for passenger flow management and has strong engineering application value.
周继彪[23]和彭羽飞[24]等提出城轨交通客流的评估因素包含客流拥挤程度、拥挤持续时间等。然而在构建铁路客站客流安全等级时,还需考虑关键区域的客流通行情况。当车站人行设施出现客流拥挤、速度降低、走行受限、短时客流剧增、密度增大且拥挤持续时间较长时,将提升区域客流风险,极易引发安全事故。为此,所提客流安全评估模型综合考虑客流的细粒度与粗粒度特征,将铁路客站客流安全评估转化为图模型[25]的辨识问题。铁路客站可抽象为有向图,其中:V为节点集(key),由车站中Nkey个关键区域组成;E为节点的边集; A 为邻接矩阵,,表征节点间的空间连接关系。图 G 中每个节点包含P个特征,每个节点在t时刻的特征向量为(i=1,2,…,N),。定义t时刻客站图中全部节点的特征向量,,则一定时间段τ内车站的安全风险为
帧率(Frames Per Second,FPS)是衡量模型处理速度的核心指标,表示模型每秒可处理的图像帧数,其值越高表明模型实时性越好。平均绝对误差(Mean Absolute Error,MAE)为预测值与真实值间绝对误差的平均值,用于表征预测的整体偏差水平。均方误差(Mean Squared Error,MSE)为预测值与真实值间平方误差的平均值,可放大较大误差的影响,对异常值更为敏感;均方根误差(Root Mean Squared Error,RMSE)为均方误差的平方根,同样适用于回归模型评价,既保留了MSE对大误差的放大特性,又因单位与原始数据一致更便于直观解释。
XIONGHao. Research on Urban Rail Transit Passenger Flow Characteristics and Weighted Network Invulnerability Based on AFC Data [D]. Xi’an: Chang’an University, 2020. in Chinese
TANGKun, WEIXianshuang, WANGShuangbao. Design of Supervisory Control System for Passenger Flow Volume Based on Infrared Flow Sensor [C]// The 9th National Conference on Photoelectronic Technology (Volume 2). Beijing: The Chinese Society of Astronautics Technical Committee on Optoelectronic Technology, 2010: 346-349. in Chinese)
[5]
KHANS D. Congestion Detection in Pedestrian Crowds Using Oscillation in Motion Trajectories [J]. Engineering Applications of Artificial Intelligence, 2019, 85: 429-443.
[6]
FUKUZAKIY, MOCHIZUKIM, MURAOK, et al. A Pedestrian Flow Analysis System Using Wi-Fi Packet Sensors to a Real Environment [C]// Proceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing: Adjunct Publication. Seattle: ACM, 2014: 721⁃730.
YUQihui, YANGHaihong, LIAosong. Research on Urban Rail Transit Passenger Flow Monitoring System Based on Multi-Source Monitoring Data Fusion [J]. Communication & Shipping, 2021, 8 (3): 68-72. in Chinese
[9]
HARIYONOJ, HOANGV D, JOK H. Motion Segmentation Using Optical Flow for Pedestrian Detection from Moving Vehicle [C]// Computational Collective Intelligence. Technologies and Applications. Cham: Springer, 2014: 204-213.
WANGAili, DONGBaotian, WANGZesheng, et al. Adaptive Background Updating Based on Optical Flow and Background Difference [J]. China Railway Science, 2014, 35 (6): 131-137. in Chinese
[12]
GIRSHICKR, DONAHUEJ, DARRELLT, et al. Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation [C]// 2014 IEEE Conference on Computer Vision and Pattern Recognition.New York: IEEE, 2014: 580-587.
[13]
REDMONJ, DIVVALAS, GIRSHICKR, et al. You Only Look Once: Unified, Real-Time Object Detection [C]// 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE, 2016: 779-788.
[14]
LIUQ L, GUOQ, WANGW, et al. An Automatic Detection Algorithm of Metro Passenger Boarding and Alighting Based on Deep Learning and Optical Flow [J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-13.
[15]
ZHANGJ, LIUJ Z, WANGZ Z. Convolutional Neural Network for Crowd Counting on Metro Platforms [J]. Symmetry, 2021, 13 (4): 703.
[16]
SIPETASC, KEKLIKOGLOUA, GONZALESE J. Estimation of Left behind Subway Passengers through Archived Data and Video Image Processing [J]. Transportation Research Part C: Emerging Technologies, 2020, 118: 102727.
[17]
WOJKEN, BEWLEYA, PAULUSD. Simple Online and Realtime Tracking with a Deep Association Metric [C]// 2017 IEEE International Conference on Image Processing (ICIP). New York: IEEE, 2017: 3645-3649.
[18]
IDREESH, TAYYABM, ATHREYK, et al. Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds [C]// Computer Vision ECCV 2018. Cham: Springer, 2018: 544-559.
[19]
ABOUSAMRAS, HOAIM, SAMARASD, et al. Localization in the Crowd with Topological Constraints [J]. Proceedings of AAAI Conference on Artificial Intelligence, 2021, 35 (2): 872-881.
LIRui, LIPing, DAIMingrui, et al. Crowd Counting Estimation Model Based on Multi-View Projection Fusion of Railway Passenger Station Video [J]. China Railway Science, 2022, 43 (3): 182-192. in Chinese
DOUFei, PANXiaojun, QINYong, et al. Identification Method for Passenger Inflow Control in Urban Rail Transit Station Based on Cloud Model [J]. Journal of Southeast University (Natural Science Edition), 2016, 46 (6): 1318-1322. in Chinese
ZHAOBaofeng, ZOUXiaolei, QUXiaoyi, et al. Classification of Early Warning for Passenger Flow Retention at Urban Rail Transit Station Based on Simulation [J]. Urban Mass Transit, 2017, 20 (9): 107-110, 115. in Chinese
LIDewei, YINHaodong. Real-Time Forecast of Passenger Crowd Index on Urban Rail Transit Station Platform Based on Test Data of Internet of Things [J]. Journal of the China Railway Society, 2014, 36 (3): 9-13. in Chinese
[28]
胡国林.城市轨道交通线网客流风险评估方法及系统研究[D].北京:北京交通大学,2022.
[29]
HUGuolin. Research on Passenger Flow Risk Assessment Method and System of Urban Rail Transit Network [D]. Beijing: Beijing Jiaotong University, 2022. in Chinese
[30]
HENRIQUESJ F, CASEIROR, MARTINSP, et al. High-Speed Tracking with Kernelized Correlation Filters [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37 (3): 583-596.
[31]
LIANGD, XUW, BAIX. An End-to-End Transformer Model for Crowd Localization [C]// Computer Vision-ECCV 2022. Cham: Springer, 2022: 38-54.
ZHOUJibiao, ZHAOPengfei, DONGSheng, et al. Pedestrian Congestion Levels at Subway Stations with Ant Colony Algorithm [J]. Urban Transport of China, 2019, 17 (4): 105-113. in Chinese
[34]
彭羽飞.城市轨道交通客流实时感知估计研究[D].北京:北京交通大学,2023.
[35]
PENGYufei. Research on Real-time Perception Estimation of Urban Rail Transit Passenger Flow [D]. Beijing: Beijing Jiaotong University, 2023. in Chinese
[36]
GUOS, LINY, FENGN, et al. Attention Based Spatial-Temporal Graph Convolutional Networks for Traffic Flow Forecasting [J]. Proceedings of the AAAI Conference on Artificial Intelligence, 2019, 33 (1): 922-929.
[37]
CHUNGJ, GÜLÇEHREC, CHOK, et al. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling [J]. ArXiv e-Prints, 2014: arXiv:1412.3555 [cs. CV].