Machine Learning-based Simulation Research of On-line Subway Pedestrian Flow Control

doi:10.16182/j.issn1004731x.joss.21-0835

Abstract

Abstract:

For the online optimization of pedestrian flow control in subway station, an algorithm frame for pedestrian flow control in subway station based on machine learning is designed. The pedestrian flow control process of a subway station during morning rush hour is selected,and the agent-based model is built to simulate the control process. The training data is collected through the multiple runs of the model, which is used as the input of deep reinforcement learning network, and the mature net is obtained through adequate training to provide the optimizing scheduling policy. Linking the actual data with the mature net to realize the real-time schedule optimization of subway pedestrian flow control. Simulation experiments show that the framework of the deep reinforcement learning can realize the on-line optimization and the performance is better than traditional algorithm.

Key words: deep reinforcement learning, pedestrian flow control, scheduling strategy in subway station, online simulation, real-time optimization

CLC Number:

TP391.9

Jiajie Shi, Peng Yang, Yannan Pi. Machine Learning-based Simulation Research of On-line Subway Pedestrian Flow Control[J]. Journal of System Simulation, 2023, 35(2): 386-395.

Figures/Tables 18

Fig. 1

Fig. 2

Fig. 3

Table 1

Table 2

Table 3

Table 4

Weight values of each factor

权重系数	通过一致性检验的权重
$α 1$	0.1
$α 2$	0.3
$α 3$	0.2
$α 4$	0.4

Table 4

Fig. 4

Fig. 5

Fig. 6

Table 5

Table 6

Table 7

Table 8

Fig. 7

Fig. 8

Fig. 9

Fig. 10

References 14

1	王媛媛. 基于行人动力学的地铁车站客流仿真模型[J].电子测试, 2018(12): 49-50, 48.
	Wang Yuanyuan. Passenger Flow Simulation Model for Metro Station Based on Pedestrian Dynamics[J]. Electrical Test, 2018(12): 49-50, 48.
2	李建华, 陈伟, 陈祥儒. 基于AnyLogic的人流增多地铁站系统再优化[J]. 科学技术与工程, 2020, 20(33): 13847-13851.
	Li Jianhua, Chen Wei, Chen Xiangru. Simulation and Optimization of Metro Station System Based on AnyLogic[J]. Science Technology and Engineering, 2020, 20(33): 13847-13851.
3	鲁晨, 孙健, 杨涛. 综合交通枢纽客流行人换乘组织优化及仿真[J]. 系统工程, 2020, 38(3): 84-91.
	Lu Chen, Sun Jian, Yang Tao. Optimum Design and Simulation of Pedestrian Passenger Transfer Operation in Integrated Transportation Hub[J]. Systems Engineering, 2020, 38(3): 84-91.
4	杨天阳, 朱志国. 基于AnyLogic的地铁车站通道设施设备规模与布局分析[J]. 交通运输工程与信息学报, 2017, 15(1): 115-121.
	Yang Tianyang, Zhu Zhiguo. Scale and Layout Analyses of Subway Station Facilities Based on AnyLogic Software[J]. Journal of Transportation Engineering and Information, 2017, 15(1): 115-121.
5	Caliendo Ciro, Ciambelli Paolo, Rossella Del Regno, et al. Modelling and Numerical Simulation of Pedestrian Flow Evacuation from a Multi-Storey Historical Building in the Event of Fire Applying Safety Engineering Tools[J]. Journal of Cultural Heritage(S1296-2074), 2020, 41: 188-199.
6	龙顺忠, 王志建, 刘皓. 网联环境下基于深度强化学习的单路口交通信号控制优化[J]. 工业控制计算机, 2020, 33(10): 16-19, 22.
	Long Shunzhong, Wang Zhijian, Liu Hao. Based on Deep Reinforcement Learning Optimization of Traffic Signal Control at Single Intersection[J]. Industrial Control Computer, 2020, 33(10): 16-19, 22.
7	Huang Yicai, Chen Jiayuan, Duan Qiannan, et al. A Fast Antibiotic Detection Method for Simplified Pretreatment Through Spectra-Based Machine Learning[J]. Frontiers of Environmental Science & Engineering(S2095-2201), 2022, 16(3): 135-146.
8	Abdelaali Hassaine, Salimi Khorshidi Gholamreza, Dexter Canoy, et al. Untangling the Complexity of Multimorbidity with Machine Learning[J]. Mechanisms of Ageing and Development(S0047-6374), 2020, 190: 111325.
9	张琪, 曾俊杰, 许凯, 等. 基于机器学习的计算机生成兵力行为建模研究综述[J]. 系统仿真学报, 2021, 33(2): 280-287.
	Zhang Qi, Zeng Junjie, Xu Kai, et al. Behavior Modeling for Computer Gederated Forces Based on Machine Learning[J]. Journal of System Simulation, 2021, 33(2): 280-287.
10	张伟罡, 龚希武. 基于机器学习的工业炉智能控制与实现[J]. 工业炉, 2020, 42(3): 45-49.
	Zhang Weigang, Gong Xiwu. Intelligent Control and Implementation of Industrial Furnace Based on Machine Learning[J]. Industrial Furnace, 2020, 42(3): 45-49.
11	李增有. 基于机器学习的雷达目标跟踪算法研究[J]. 信息技术与信息化, 2020(8): 217-219.
	Li Zengyou. Research on Radar Target Tracking Algorithm Based on Machine Learning[J]. Information Technology and Informatization, 2020(8): 217-219.
12	张荣霞, 武长旭, 孙同超, 等. 深度强化学习及在路径规划中的研究进展[J].计算机工程与应用, 2021, 57(19): 44-56.
	Zhang Rongxia, Wu Changxu, Sun Tongchao, et al. Progress on Deep Reinforcement Learning in Path Planning[J]. Computer Engineering and Applications, 2021, 57(19): 44-56.
13	Wang Haonan, Liu Ning, Zhang Yiyun, et al. Deep Reinforcement Learning: a Survey[J]. Frontiers of Information Technology & Electronic Engineering(S2095-9184), 2020, 21(12): 1726-1744.
14	Neeraj R R, Nithin R P, Niranjhan P, et al. Modelling and Simulation of Discrete Manufacturing Industry[J]. Materials Today: Proceedings(S2214-7853), 2018, 5(11): 24971-24983.

时段	该时段首发时间	时间间隔
07:00—08:00	07:01	每4 min一趟
08:00—09:00	08:00	每3 min一趟
09:00—10:00	09:01	每3 min一趟

参数	初值
安检通道数量/个	16
人工安检数量/个	4
闸机数量/个	84
隔离栏长度/m	20
电梯数量/个	24

参数	初值
安检速度/(个/m)	Triangular(2.5, 5.5, 2.5)
行人行走速度/(m/s)	Triangular(0.8, 1.6, 0.8)

输入	变量取值
当前时刻观测变量	站外限行区行人流密度
	安检区行人流密度
	闸机通行行人流密度
	站台行人流密度
上一时刻控制变量	安检通道数量
	人工安检通道数量
	闸机数量
	电动步梯数量

输出	变量取值
下一时刻控制变量	安检通道数量
	人工安检通道数量
	闸机数量
	电动步梯数量