Prediction of the Metro Section Passenger Flow Based on Time-Space Characteristic

Abstract:

At present, the real-time data of the section passenger flow can’t be acquired in the process of urban rail transit operation, which brings some difficulties to the passenger monitoring and controlling. After analyzing the spatio-temporal complexity characters, this paper adopted BP neural network to predict the section passenger flow, which was based on the first three consecutive periods of the in and out station traffic and the next period of the section passenger flow data. Finally, Fuxingmen and Xidan in Beijing urban rail transit were selected, their first three consecutive periods of the in and out traffic are put as the input data, and the next period of the section traffic as the output data, then BP neural network was used to train and predict under MATLAB. The anticipant and the actual output results are well fitted, which proves that the data processing method is effective and the parameters of the BP neural network are reasonable, and it can provide theoretical reference for the transport operators to some extent.

• Periodical: Applied Mechanics and Materials (Volumes 397-400)
• Pages: 1038-1044
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.397-400.1038
• Online since: September 2013
• Authors: Jun Jin, Yan Hui Wang, Man Li
• Keywords: BP Neural Network (BPNN), Forecasting, Section Passenger Flow, Time-Space Character

References

[1] Shi Junqing, Cheng Li: WIT Transactions on the Built Environment, Vol. 140(2013), pp.177-181.

[2] Zhou Jing，Lam William H K，Heydecker Benjamin G: Transportation Research Part B: Methodological. Vol. 39 (2005)，pp.519-544.

[3] Tao Zhixiang，Zhang Ning，Du Bo: Urban public transport. Vol. 2 (2004)，pp.33-35(in Chinese).

[4] Pei Yulong, Wang Xiaoning: Journal of Harbin Institute of Technology, Vol. 8 (2004)，p.012.

[5] Wang Jin，Shi Qixin: Journal of Henan University of Science &Technology (natural science). (2005)，pp.1-3(in Chinese).

[6] Wang Yanhui，Jia Linmin，Wang Zhuo，et al: China Railway Science. Vol. 26 (2005), pp.130-135(in Chinese).

[7] Zhang Defeng，et al. MATLAB neural network application design [M]. Beijing: Mechanical Industry Press，2009(in Chinese).