Passenger Flow Prediction of Scenic Spot Using a GCN–RNN Model

Zhijie Xu,Liyan Hou,Jianqin Zhang,Yueying Zhang

doi:10.3390/su14063295

Zhijie Xu, Liyan Hou + Show 2 more

Open Access

https://doi.org/10.3390/su14063295

Copy DOI

Abstract

The prediction and control of passenger flow in scenic spots is very important to the traffic management and safety of scenic spots. This study aims to predict the passenger flow of a scenic spot based on the passenger flow of the bus and subway stations around the scenic spots. We propose a passenger flow prediction model based on graph convolutional network–recurrent neural network (GCN–RNN). First, a “graph” is constructed according to the geographical relationship between the scenic spot and the surrounding bus and subway stations. Then, characteristics of surrounding areas of bus and subway stations are constructed based on the crowd behavior analysis, and these are then used as the node-information of the “graph”. Last, the GCN–RNN model is used to extract the temporal and spatial characteristics of the passenger flow data of the scenic spot to realize the prediction. The experimental results show that the proposed model is effective in passenger flow prediction in scenic spots.

Highlights

Introduction of Scenic Spot Using agraph convolutional network–recurrent neural network (GCN–RNN)A scenic spot is a place which often experiences a dense flow of people, and the passenger flow of a scenic spot refers to the number of tourists in the scenic spot at a specific time
More scholars use deep learning methods to predict passenger flow in scenic spots, including LSTM [11], improved BP neural network optimized by genetic algorithm [12], improved BP neural network optimized by particle swarm optimization [13], generalized deep belief network particle swarm optimization prediction model [14], stacked autoencoders with deep neural networks [15], and adaptive networks [16], etc
We compare the model proposed in this paper with the time series models RNN and LSTM used for passenger flow prediction in scenic spots

Summary

Introduction

Introduction of Scenic Spot Using aGCN–RNNA scenic spot is a place which often experiences a dense flow of people, and the passenger flow of a scenic spot refers to the number of tourists in the scenic spot at a specific time. The prediction of passenger flow in scenic spots can provide an early warning of accidents, ensure the safety of the people, reduce economic losses, and play a very important role for decision makers and the public. The first type is the traditional prediction methods of passenger flow in scenic spot, which mainly include wavelet analysis [2], EMD [3], the use of regression analysis [4], grey model [5], wavelet transform, LS-SVM [6], probability tree [7], wavelet-SVM [8], Markov, and other methods [9,10]. Another type of prediction methods is based on deep learning. More scholars use deep learning methods to predict passenger flow in scenic spots, including LSTM [11], improved BP neural network optimized by genetic algorithm [12], improved BP neural network optimized by particle swarm optimization [13], generalized deep belief network particle swarm optimization prediction model [14], stacked autoencoders with deep neural networks [15], and adaptive networks [16], etc

Objectives

Methods

Results

Discussion

Conclusion