Traffic flow prediction based on BILSTM model and data denoising scheme
Zhong-Yu Li(李中昱)1,2,3, Hong-Xia Ge(葛红霞)1,2,3,†, and Rong-Jun Cheng(程荣军)1,2,3
1 Faculty of Maritime and Transportation, Ningbo University, Ningbo 315211, China; 2 Jiangsu Provincial Collaborative Innovation Center for Modern Urban Traffic Technologies, Nanjing 210096, China; 3 National Traffic Management Engineering and Technology Research Center Ningbo University Subcenter, Ningbo 315211, China
Abstract Accurate prediction of road traffic flow is a significant part in the intelligent transportation systems. Accurate prediction can alleviate traffic congestion, and reduce environmental pollution. For the management department, it can make effective use of road resources. For individuals, it can help people plan their own travel paths, avoid congestion, and save time. Owing to complex factors on the road, such as damage to the detector and disturbances from environment, the measured traffic volume can contain noise. Reducing the influence of noise on traffic flow prediction is a piece of very important work. Therefore, in this paper we propose a combination algorithm of denoising and BILSTM to effectively improve the performance of traffic flow prediction. At the same time, three denoising algorithms are compared to find the best combination mode. In this paper, the wavelet (WL) denoising scheme, the empirical mode decomposition (EMD) denoising scheme, and the ensemble empirical mode decomposition (EEMD) denoising scheme are all introduced to suppress outliers in traffic flow data. In addition, we combine the denoising schemes with bidirectional long short-term memory (BILSTM) network to predict the traffic flow. The data in this paper are cited from performance measurement system (PeMS). We choose three kinds of road data (mainline, off ramp, on ramp) to predict traffic flow. The results for mainline show that data denoising can improve prediction accuracy. Moreover, prediction accuracy of BILSTM+EEMD scheme is the highest in the three methods (BILSTM+WL, BILSTM+EMD, BILSTM+EEMD). The results for off ramp and on ramp show the same performance as the results for mainline. It is indicated that this model is suitable for different road sections and long-term prediction.
Fund: Project supported by the Program of Humanities and Social Science of the Education Ministry of China (Grant No. 20YJA630008), the Natural Science Foundation of Zhejiang Province, China (Grant No. LY20G010004), and the K C Wong Magna Fund in Ningbo University, China.
Corresponding Authors:
Hong-Xia Ge
E-mail: gehongxia@nbu.edu.cn
Cite this article:
Zhong-Yu Li(李中昱), Hong-Xia Ge(葛红霞), and Rong-Jun Cheng(程荣军) Traffic flow prediction based on BILSTM model and data denoising scheme 2022 Chin. Phys. B 31 040502
[1] Do L N N, Vu H L, Vo B Q, Liu Z Y and Phung D 2019 Transportation Research Part C:Emerging Technologies108 12 [2] Wu Y, Tan H, Qin L, Ran B and Jiang Z X 2018 Transportation Research Part C:Emerging Technologies90 166 [3] Williams B M 2001 Transportation Research Record1776 194 [4] Chen C Y, Hu J M, Meng Q and Zhang Y 2011 2011 IEEE Intelligent Vehicles Symposium (IV) pp. 607-612 [5] Kumar S V and Vanajakshi L 2015 European Transport Research Review7 1 [6] Giraka O and Selvaraj V K 2020 Transportation Letters12 483 [7] Chen R, Liang C Y, Hong W C and Gu D X 2015 Applied Soft Computing26 435 [8] Zhang L, Alharbe N R, Luo G C, Yao Z Y and Li Y 2018 Tsinghua Science and Technology23 479 [9] Cheng A Y, Jiang X, Li Y F, Zhang C and Zhu H 2017 Physica A466 422 [10] Li C, Wang X D, Cheng Z W and Bai Y 2020 IEEE Access8 19717 [11] More R, Mugal A, Rajgure S, Adhao R B and Pachghare V K 2016 2016 International Conference on Computing, Analytics and Security Trends (CAST), pp. 52-57 [12] Kumar K, Parida M and Katiyar V K 2015 Transport30 397 [13] Zhang W B, Yu Y H, Qi Y, Shu F and Wang Y H 2019 Transportmetrica A:Transport Science15 1688 [14] Yu F, Wei D, Zhang S T and Shao Y L 2019 2019 4th International Conference on Intelligent Transportation Engineering (ICITE), pp. 99-103 [15] An J Y, Fu L, Hu M, Chen W H and Zhan J W 2019 IEEE Access7 20708 [16] Sha S W, Li J, Zhang K, Yang Z F, Wei Z J, Li X Y and Zhu X 2020 IEEE Access8 15232 [17] Duan Y J, Yisheng L V and Wang F Y 2016 IEEE 19th international conference on intelligent transportation systems (ITSC), pp. 1053-1058 [18] Wu Y K, Tan H C, Qin L Q, Ran B and Jiang Z X 2018 Transportation Research Part C:Emerging Technologies90 166 [19] Hou Q Z, Leng J Q, Ma G S, Liu W Y and Cheng Y X 2019 Physica A527 121065 [20] Tang J J, Zhang G H, Wang Y H, Wang H and Liu Fang 2015 Transportation Research Part C:Emerging Technologies51 29 [21] Zheng H F, Lin F, Feng X X and Chen Y J 2020 IEEE Transactions on Intelligent Transportation Systems22 6910 [22] Chen X Q, Li Z B, Yang Y S, Qi L and Ke R 2020 IEEE Transactions on Intelligent Transportation Systems22 3190 [23] Chen X Y, He Z C and Sun L J 2019 Transportation research part C:emerging technologies98 73 [24] Zheng T X, Girgis A A and Makram E B 2000 Electric Power Systems Research54 11 [25] Xie Y C, Zhang Y L and Ye Z R 2007 Computer-Aided Civil and Infrastructure Engineering22 326 [26] Dunne S and Ghosh B 2013 IEEE Transactions on Intelligent Transportation Systems14 370 [27] Xiao H, Sun H Y, Ran B and Oh Y 2003 Transportation Research Record:Journal of the Transportation Research Board1836 16 [28] Yang W, Yang D Y, Zhao Y L and Gong J L 2010 2010 International Conference on Logistics Systems and Intelligent Management (ICLSIM). 2 pp. 969-972 [29] Wang H Z, Liu L, Dong S J, Qian Z and Wei H 2016 Transportmetrica B:Transport Dynamics4 159 [30] Chen X Q, Chen H X, Yang Y S, Wu H F, Zhang W H, Zhao J S and Xiong Y 2021 Physica A565 125574 [31] Tang J J, Gao F, Liu F and Chen X Q 2020 IEEE Access8 11546 [32] Chen X Q, Lu J Q, Zhao J S, Qu Z J, Yang Y S and Xian J F 2020 Sustainability12 3678 [33] Huang D W 2003 Physica A:Statistical Mechanics and its Applications329 298 [34] Huang N E, Shen Z, Long S R, Wu M C, Shih H H, Zheng Q A, Yen N C, Tung C C and Liu H H 1998 Proc. Royal Soc. London. Series A454 903 [35] Hochreiter S and Schmidhuber J 1997 Neural Computation9 1735
Altmetric calculates a score based on the online attention an article receives. Each coloured thread in the circle represents a different type of online attention. The number in the centre is the Altmetric score. Social media and mainstream news media are the main sources that calculate the score. Reference managers such as Mendeley are also tracked but do not contribute to the score. Older articles often score higher because they have had more time to get noticed. To account for this, Altmetric has included the context data for other articles of a similar age.
