The effect of the dataset on evaluating urban traffic prediction

被引:13
作者
Hou, Yue [1 ]
Chen, Jiaxing [1 ]
Wen, Sheng [2 ]
机构
[1] Lanzhou Jiaotong Univ, Sch Elect & Informat Engn, 88 Aiming West Rd, Lanzhou 730070, Gansu, Peoples R China
[2] Swinburne Univ Technol, MIEEE Sch Software & Elect Engn, Hawthorn, Vic 3122, Australia
基金
中国国家自然科学基金;
关键词
Traffic state prediction; LSTM; GRU; SAE; Traffic data; XiAn Road Traffic; INTELLIGENT TRANSPORTATION SYSTEMS; TRAVEL-TIME PREDICTION; REAL-TIME; TRAJECTORY RECONSTRUCTION; NEURAL-NETWORK; TERM; STATE; COLLECTION; VOLUME; MODEL;
D O I
10.1016/j.aej.2020.09.038
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
With the continuous development of economic strength and science and technology, the construction of Intelligent Transportation System(ITS) has become a new development direction in many cities. A complete and accurate traffic dataset can improve the accuracy of traffic prediction and promote the construction of ITS in cities. Most of the existing traffic datasets are collected on highways, and they are only one-way road data. There is little analysis of the impact of weather on traffic prediction, and more traffic auxiliary information is lacking at the same time. The use of such datasets for experiments can lead to inaccurate and unconvincing results, which is of little significance for the study of urban road prediction reference. In this paper, we are motivated to develop a new dataset for the evaluation of Metropolitan Traffic Prediction. Our dataset(XiAn Road Traffic) collected 308 urban road data and included two-way road data, weather data, driving angles, and congestion levels. XiAn Road Traffic can provide help for urban road state prediction and intelligent transportation city construction. We use the current more popular machine learning model for experiments. It is also proved by experiments that our dataset is more accurate and persuasive than the prediction results of other datasets. (C) 2020 The Authors. Published by Elsevier B.V. on behalf of Faculty of Engineering, Alexandria University.
引用
收藏
页码:597 / 613
页数:17
相关论文
共 71 条
[1]  
[Anonymous], 2018, IEEE ACCESS
[2]  
[Anonymous], IEEE T WIRELESS COMM
[3]   Spatiotemporal Patterns in Large-Scale Traffic Speed Prediction [J].
Asif, Muhammad Tayyab ;
Dauwels, Justin ;
Goh, Chong Yang ;
Oran, Ali ;
Fathi, Esmail ;
Xu, Muye ;
Dhanya, Menoth Mohan ;
Mitrovic, Nikola ;
Jaillet, Patrick .
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2014, 15 (02) :794-804
[4]   Performance Evaluation of Travel-Time Estimation Methods for Real-Time Traffic Applications [J].
Ban, Xuegang Jeff ;
Li, Yuwei ;
Skabardonis, Alexander ;
Margulici, J. D. .
JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS, 2010, 14 (02) :54-67
[5]  
Cai Kuanqi, SINGULAR POINT PROBA
[6]   Cross-Correlation Analysis and Multivariate Prediction of Spatial Time Series of Freeway Traffic Speeds [J].
Chandra, Srinivasa Ravi ;
Al-Deek, Haltham .
TRANSPORTATION RESEARCH RECORD, 2008, (2061) :64-76
[7]   Use of sequential learning for short-term traffic flow forecasting [J].
Chen, H ;
Grant-Muller, S .
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES, 2001, 9 (05) :319-336
[8]   Applied Research on Traffic Information Collection Based on Wireless Sensor Networks [J].
Chen, Xiaoquan ;
Zhang, Jihong ;
Qian, Shao ;
Xu, Peng .
2012 INTERNATIONAL CONFERENCE ON FUTURE ELECTRICAL POWER AND ENERGY SYSTEM, PT A, 2012, 17 :602-606
[9]   Predicting travel times for the South Jersey real-time motorist information system [J].
Chien, SIJ ;
Liu, XB ;
Ozbay, K .
TRANSPORTATION DATA RESEARCH: PLANNING AND ADMINISTRATION, 2003, (1855) :32-40
[10]   Dynamic travel time prediction with real-time and historic data [J].
Chien, SIJ ;
Kuchipudi, CM .
JOURNAL OF TRANSPORTATION ENGINEERING, 2003, 129 (06) :608-616