Deep Learning Model for Traffic Speed Prediction Considering Spatial-Temporal Correlation
Author:
- 31 March 2022
- journal article
- research article
- Published by Korean Institute of Information Technology in The Journal of Korean Institute of Information Technology
- Vol. 20 (3) , 19-26
Abstract
교통량과 속도는 지능형 교통 시스템을 구축하기 위해 필요한 가장 중요한 교통 정보이다. 최근 사물인터넷, 빅데이터, 인공지능 등의 기술 발전에 따라 다양한 딥러닝 기술들이 교통 정보 예측에 많이 활용되고 있다. 본 논문에서는 도로의 공간적 특징과 시간에 따른 속도 변화 특징을 반영하기 위하여 3개의 모델을 결합한 CNN-RNN-Attention 속도 예측 모델을 제안한다. 컨볼루션 신경망(CNN)과 순환신경망(RNN)을 활용하여 도로의 공간적 특성과 시계열 특성을 각각 학습하고, 학습된 결과에 어텐션(Attention) 기법을 적용하여 가중치를 부여함으로써 성능을 향상시켰다. 제안된 모델은 한국도로공사 링크 통행속도 데이터를 활용하여 실험을 진행하였으며, CNN 단일 모델, GRU 단일 모델 그리고 CNN과 GRU을 결합한 모델 대비 성능이 우수함을 증명하였다.Keywords
Funding Information
- Dongyang Mirae University
This publication has 18 references indexed in Scilit:
- A novel work zone short-term vehicle-type specific traffic speed prediction model through the hybrid EMD–ARIMA frameworkTransportmetrica B: Transport Dynamics, 2015
- Long short-term memory neural network for traffic speed prediction using remote microwave sensor dataTransportation Research Part C: Emerging Technologies, 2015
- Statistical methods versus neural networks in transportation research: Differences, similarities and some insightsTransportation Research Part C: Emerging Technologies, 2011
- Forecasting urban traffic flow by SVR with continuous ACOApplied Mathematical Modelling, 2011
- Online-SVR for short-term traffic flow prediction under typical and atypical traffic conditionsExpert Systems with Applications, 2009
- Travel-Time Prediction With Support Vector RegressionIEEE Transactions on Intelligent Transportation Systems, 2004
- Modeling and Forecasting Vehicular Traffic Flow as a Seasonal ARIMA Process: Theoretical Basis and Empirical ResultsJournal of Transportation Engineering, 2003
- Traffic Prediction Using Multivariate Nonparametric RegressionJournal of Transportation Engineering, 2003
- Comparison of parametric and nonparametric models for traffic flow forecastingTransportation Research Part C: Emerging Technologies, 2002
- Multivariate Vehicular Traffic Flow Prediction: Evaluation of ARIMAX ModelingTransportation Research Record: Journal of the Transportation Research Board, 2001