An Improved Bayesian Combination Model for Short-Term Traffic Prediction With Deep Learning

被引:133
作者
Gu, Yuanli [1 ]
Lu, Wenqi [2 ]
Xu, Xinyue [3 ]
Qin, Lingqiao [4 ]
Shao, Zhuangzhuang [1 ]
Zhang, Hanyu [3 ]
机构
[1] Beijing Jiaotong Univ, Key Lab Transport Ind Big Data Applicat Technol C, Minist Transport, Beijing 100044, Peoples R China
[2] Southeast Univ, Sch Transportat, Nanjing 211189, Peoples R China
[3] Beijing Jiaotong Univ, State Key Lab Rail Traff Control & Safety, Beijing 100044, Peoples R China
[4] Univ Wisconsin, Dept Civil & Environm Engn, Traff Operat & Safety Lab, Madison, WI 53706 USA
基金
中国国家自然科学基金;
关键词
Predictive models; Deep learning; Correlation; Neural networks; Bayes methods; Data models; Roads; Urban road; short-term traffic flow prediction; improved Bayesian combination method; gated recurrent unit neural network; microwave data; FLOW PREDICTION; NEURAL-NETWORK; BIG DATA; CLASSIFICATION; UNCERTAINTY; LSTM;
D O I
10.1109/TITS.2019.2939290
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Short-term traffic volume prediction, which can assist road users in choosing appropriate routes and reducing travel time cost, is a significant topic of intelligent transportation system. To overcome the error magnification phenomena of traditional combination methods and to improve prediction performance, this paper proposes an improved Bayesian combination model with deep learning (IBCM-DL) for traffic flow prediction. First, an IBCM framework is established based on the new BCM framework proposed by Wang. Then, correlation analysis is used to analyze the relevance between the historical traffic flow and the traffic flow within the current interval. Three sub-predictors including the gated recurrent unit neural network (GRUNN), autoregressive integrated moving average (ARIMA), and radial basis function neural network (RBFNN) are incorporated into the IBCM framework to take advantage of each method. The real-world traffic volume data captured by microwave sensors located on the expressways of Beijing was used to validate the proposed model in multiple scenarios. The overall results illustrate that the IBCM-DL model outperforms the other state-of-the-art methods in terms of accuracy and stability.
引用
收藏
页码:1332 / 1342
页数:11
相关论文
共 65 条
[61]   An Improved K-nearest Neighbor Model for Short-term Traffic Flow Prediction [J].
Zhang, Lun ;
Liu, Qiuchen ;
Yang, Wenchen ;
Wei, Nai ;
Dong, Decun .
INTELLIGENT AND INTEGRATED SUSTAINABLE MULTIMODAL TRANSPORTATION SYSTEMS PROCEEDINGS FROM THE 13TH COTA INTERNATIONAL CONFERENCE OF TRANSPORTATION PROFESSIONALS (CICTP2013), 2013, 96 :653-662
[62]   A hybrid short-term traffic flow forecasting method based on spectral analysis and statistical volatility model [J].
Zhang, Yanru ;
Zhang, Yunlong ;
Haghani, Ali .
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, 2014, 43 :65-78
[63]   LSTM network: a deep learning approach for short-term traffic forecast [J].
Zhao, Zheng ;
Chen, Weihai ;
Wu, Xingming ;
Chen, Peter C. Y. ;
Liu, Jingmeng .
IET INTELLIGENT TRANSPORT SYSTEMS, 2017, 11 (02) :68-75
[64]   Short-term freeway traffic flow prediction: Bayesian combined neural network approach [J].
Zheng, WZ ;
Lee, DH ;
Shi, QX .
JOURNAL OF TRANSPORTATION ENGINEERING, 2006, 132 (02) :114-121
[65]   Spectral-Spatial Residual Network for Hyperspectral Image Classification: A 3-D Deep Learning Framework [J].
Zhong, Zilong ;
Li, Jonathan ;
Luo, Zhiming ;
Chapman, Michael .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2018, 56 (02) :847-858