A Method for Traffic Flow Forecasting in a Large-Scale Road Network Using Multifeatures

Chuhao Zhou,Peiqun Lin,Yang Cheng,Xukun Lin

doi:10.7307/ptt.v33i4.3709

Chuhao Zhou, Peiqun Lin + Show 2 more

Open Access

https://doi.org/10.7307/ptt.v33i4.3709

Copy DOI

Abstract

Accurate traffic prediction on a large-scale road network is significant for traffic operations and management. In this study, we propose an equation for achieving a comprehensive and accurate prediction that effectively combines traffic data and non-traffic data. Based on that, we developed a novel prediction model, called the adaptive deep neural network (ADNN). In the ADNN, we use two long short-term memory (LSTM) networks to extract spatial-temporal characteristics and temporal characteristics, respectively. A backpropagation neural network (BPNN) is also employed to represent situations from contextual factors such as station index, forecast horizon, and weather. The experimental results show that the prediction of ADNN for different stations and different forecast horizons has high accuracy; even for one hour ahead, its performance is also satisfactory. The comparison of ADNN and several benchmark prediction models also indicates the robustness of the ADNN.

Highlights

Comprehensive and accurate traffic prediction of a large-scale road network plays an important role in traffic operations and management
We propose a novel deep neural network, an adaptive deep neural network (ADNN), which can achieve a multistep prediction for multiple locations in a road network with satisfying accuracy
Based on Equation 2, we propose a novel architecture of the neural network, namely, ADNN, which can seek an optimal state between completeness and precision

Summary

Introduction

Comprehensive and accurate traffic prediction of a large-scale road network plays an important role in traffic operations and management. The road network should be considered as a whole since the congestion of a certain road is associated with many other road sections. An accurate prediction for a large-scale road network is helpful for congestion analysis and vehicle route planning, which can help alleviate congestion through load balancing [1]. Statistical methods were often employed for traffic prediction [2,3,4]. Due to the availability of traffic big data [8, 9] and the rapid development of neural networks [1012], using spatial-temporal data for traffic prediction has become more effective [13]

Methods

Results

Conclusion