Forecast of Subway Inbound Passenger Flow based on LSTM Model and BP Neural Network
DOI:
https://doi.org/10.54097/05h04s79Keywords:
LSTM, BP, passenger flow data, root mean squared error.Abstract
Passenger flow prediction for rail transit is an extremely important research topic. This paper constructs a long-short-term memory model, and proposes a simple optimization method. At the same time this paper constructs a backpropagation neural network model and adjusts the size of the hidden layer through testing. Within 15 days, swipe card data from a randomly chosen tube station in a city is subjected to resampling, normalization, and other data preparation techniques. The processed data is then split into an 80% training set and a 20% test set. The root-mean-square error test analysis is carried out on the three types of models constructed to judge the fitting effect of the data. The conclusion is that the BP neural network model is relatively easy to build a shallow network structure, which is typically less than or equal to 3. Additionally, the data fitting effect outperforms both the LSTM model and the straightforward optimization approach. Especially for passenger flow fluctuations in a short period of time, the BP neural network can have good prediction ability.
Downloads
References
Guoguang He, Yu Li, Shoufeng Ma. Discussion on short-term Traffic flow prediction Method based on Mathematical model. Systems Engineering Theory & Practice, 2000, (12): 51-56.
Donglong Yang. Research on short-term road traffic flow prediction based on ARIMA model. Electronic design engineering, 2021, 29(13): 38-42.
Zhendong Xie, Xueqin Liu, Jincheng Wu, et al. Research on passenger flow prediction Model based on public transportation IC card data. Journal of Guangdong University of Technology, 2018, 35(01): 16-22.
Dongyang Chen, et al. Research on Metro Passenger Travel Time Prediction Based on Passenger Flow Big Data Analysis and Support Vector Regression. Modern Urban Rail Transit, 2020, (09): 70-76.
Guo Wen, Xiao Weizhou, Qin Feifei. Passenger flow prediction of Metro station based on Support vector Machine model. Hebei Industrial Science and Technology, 2019, 36(01): 31-35.
Xu Wenxing, Zhang Yu. Short-term Passenger flow prediction of orbital stations based on Random forest algorithm. Journal of Wuhan University of Technology, 2012, 46(03): 406-410.
Qian Huimin, et al. Sports events surrounding rail site based on AFC data of short-term passenger flow prediction. Journal of transportation engineering, 2020, 20 (6): 30-36.
Zhang Yawei, Chen Ruifeng, LIU Xiaoyan. Door control loop of the neural network based passenger flow prediction. Journal of railway transportation and economy, 2022, 44(9): 96-102.
Yang Liu. Application of Deep Learning in Transportation. Information Technology and Informatization, 2022, 263(02): 192-195.
Zuo Tong, et al. Passenger flow prediction of urban rail transit stations based on deep fusion of spatio-temporal characteristics. Comprehensive Transportation, 2023, 45(05): 67-74.
Feng Biyu. Research on short-term passenger flow prediction of urban rail transit based on CNN-LSTM combined model. East China jiaotong university, 2020.
Zeng Lu, et al. Research on short-term Passenger Flow Prediction of urban rail Transit Based on CEEMDAN-IPSO-LSTM. Journal of railway science and engineering, 2023, 1-14.
Li Gao-sheng, Peng Ling, LI Xiang, Wu Tong. Research on short-term passenger flow prediction of urban bus station based on LSTM. Journal of Highway and Transportation Technology, 2019, 36(2): 128-135.
Downloads
Published
Issue
Section
License
Copyright (c) 2023 Highlights in Science, Engineering and Technology
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
