Incorporating Multivariate Auxiliary Information for Traffic Prediction on Highways
Abstract
:1. Introduction
- We propose a novel BiLSTM-based model, which combines multivariate auxiliary information in highways to learn the representation of features for better prediction performance.
- We design a multi-horizon strategy and use the soft attention module to integrate the different effects of different horizons.
- We conduct comprehensive experiments on two datasets, and the results show that the proposed model MMLSTM achieves better performance than the baselines.
2. Related Work
2.1. Traditional Methods for Time Series Prediction
2.2. DNN for Traffic Flow Prediction
2.3. Attention for Traffic Flow Prediction
3. Methodology
3.1. Traffic Flow Prediction Problem
3.2. MMLSTM
3.2.1. Embedding and Fusion Layer
3.2.2. Multi-Horizon BiLSTM Layer and Attention Layer
3.2.3. Traffic Flow Prediction Layer
4. Experiments
4.1. Datasets
4.2. Evaluation Metrics
4.3. Comparison Methods
- HA [34] Historical Average method forecasts the average value of the same time in the training data.
- SVR SVR is a machine learning framework based on logistic regression.
- XGBoost [35] XGBoost is a parallel regression tree model combined with boosting.
- LightGBM [36] LightGBM is a highly efficient gradient boosting decision tree.
- GRU [37] GRU introduces a gating mechanism in RNN.
- LSTM [38] Long Short-Term Memory network is an improved RNN-based model, which contains three gates to preserve the long dependencies in sequence.
- SCGRU [39] A Sparse-Connection GRU model (SCGRU) focuses on reducing the storage and computation costs using a controllable threshold on the absolute value of the pre-trained GRU weights.
- ST-Norm [40] ST-Norm contains two normalization modules that refine the local and high-frequency components of raw data and can be integrated into deep learning models such as Wavenet and Transformer.
4.4. Implementation and Settings
4.5. Experiment Results
- RQ1: Does MMLSTM outperform baseline methods in four traffic flow prediction tasks?
- RQ2: How do the value setting of parameters affect the MMLSTM’s performance?
- RQ3: Do different components of the MMLSTM improve the performance?
4.5.1. Analysis of Comparative Experiment Results (RQ1)
4.5.2. Analysis of Hyperparameter Experiment Results (RQ2)
4.5.3. Analysis of Ablation Experiment Results (RQ3)
4.5.4. Analysis of Case Study
5. Conclusions and Future Work
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Fang, S.; Zhang, Q.; Meng, G.; Xiang, S.; Pan, C. GSTNet: Global spatial-temporal network for traffic flow prediction. In Proceedings of the IJCAI, Macao, China, 10–16 August 2019; pp. 2286–2293. [Google Scholar]
- Ren, Y.; Zhao, D.; Luo, D.; Ma, H.; Duan, P. Global-local temporal convolutional network for traffic flow prediction. IEEE Trans. Intell. Transp. Syst. 2020, 23, 1578–1584. [Google Scholar] [CrossRef]
- Pan, B.; Demiryurek, U.; Shahabi, C. Utilizing real-world transportation data for accurate traffic prediction. In Proceedings of the 2012 IEEE 12th International Conference on Data Mining, IEEE, Brussels, Belgium, 10–13 December 2012; pp. 595–604. [Google Scholar]
- Wu, F.; Wang, H.; Li, Z. Interpreting traffic dynamics using ubiquitous urban data. In Proceedings of the 24th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, Burlingame, CA, USA, 31 October–3 November 2016; pp. 1–4. [Google Scholar]
- Zhang, J.; Zheng, Y.; Qi, D.; Li, R.; Yi, X. DNN-based prediction model for spatio-temporal data. In Proceedings of the 24th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, Burlingame, CA, USA, 31 October–3 November 2016; pp. 1–4. [Google Scholar]
- Zhang, J.; Zheng, Y.; Qi, D. Deep spatio-temporal residual networks for citywide crowd flows prediction. In Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, San Francisco, CA, USA, 4–9 February 2017. [Google Scholar]
- Yu, R.; Li, Y.; Shahabi, C.; Demiryurek, U.; Liu, Y. Deep learning: A generic approach for extreme condition traffic forecasting. In Proceedings of the 2017 SIAM International Conference on Data Mining, SIAM, Houston, TX, USA, 27–29 April 2017; pp. 777–785. [Google Scholar]
- Yao, H.; Wu, F.; Ke, J.; Tang, X.; Jia, Y.; Lu, S.; Gong, P.; Ye, J.; Li, Z. Deep multi-view spatial-temporal network for taxi demand prediction. In Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, LA, USA, 2–7 February 2018; Volume 32. [Google Scholar]
- Tao, Y.; Sun, P.; Boukerche, A. A hybrid stacked traffic volume prediction approach for a sparse road network. In Proceedings of the 2019 IEEE Symposium on Computers and Communications (ISCC), IEEE, Barcelona, Spain, 29 June–3 July 2019; pp. 1–6. [Google Scholar]
- Zhao, J.; Gao, Y.; Yang, Z.; Li, J.; Feng, Y.; Qin, Z.; Bai, Z. Truck traffic speed prediction under non-recurrent congestion: Based on optimized deep learning algorithms and GPS data. IEEE Access 2019, 7, 9116–9127. [Google Scholar] [CrossRef]
- Liu, Y.; Zheng, H.; Feng, X.; Chen, Z. Short-term traffic flow prediction with Conv-LSTM. In Proceedings of the 2017 9th International Conference on Wireless Communications and Signal Processing (WCSP), IEEE, Nanjing, China, 11–13 October 2017; pp. 1–6. [Google Scholar]
- Guo, K.; Hu, Y.; Qian, Z.; Liu, H.; Zhang, K.; Sun, Y.; Gao, J.; Yin, B. Optimized graph convolution recurrent neural network for traffic prediction. IEEE Trans. Intell. Transp. Syst. 2020, 22, 1138–1149. [Google Scholar] [CrossRef]
- Guo, S.; Lin, Y.; Wan, H.; Li, X.; Cong, G. Learning dynamics and heterogeneity of spatial-temporal graph data for traffic forecasting. IEEE Trans. Knowl. Data Eng. 2021, 34, 5415–5428. [Google Scholar] [CrossRef]
- Liu, D.; Xu, X.; Xu, W.; Zhu, B. Graph convolutional network: Traffic speed prediction fused with traffic flow data. Sensors 2021, 21, 6402. [Google Scholar] [CrossRef] [PubMed]
- Amini, M.H.; Kargarian, A.; Karabasoglu, O. ARIMA-based decoupled time series forecasting of electric vehicle charging demand for stochastic power system operation. Electr. Power Syst. Res. 2016, 140, 378–390. [Google Scholar] [CrossRef]
- Geetha, A.; Nasira, G. Time-series modelling and forecasting: Modelling of rainfall prediction using ARIMA model. Int. J. Soc. Syst. Sci. 2016, 8, 361–372. [Google Scholar] [CrossRef]
- Zhang, Y.; Roughan, M.; Willinger, W.; Qiu, L. Spatio-temporal compressive sensing and internet traffic matrices. In Proceedings of the ACM SIGCOMM 2009 Conference on Data Communication, Barcelona, Spain, 16–21 August 2009; pp. 267–278. [Google Scholar]
- Yu, H.F.; Rao, N.; Dhillon, I.S. Temporal regularized matrix factorization for high-dimensional time series prediction. In Proceedings of the 30th International Conference on Neural Information Processing Systems, Barcelona, Spain, 5–10 December 2016; Volume 29. [Google Scholar]
- Mellit, A.; Pavan, A.M.; Benghanem, M. Least squares support vector machine for short-term prediction of meteorological time series. Theor. Appl. Climatol. 2013, 111, 297–307. [Google Scholar] [CrossRef]
- Das, M.; Ghosh, S.K. A probabilistic approach for weather forecast using spatio-temporal inter-relationships among climate variables. In Proceedings of the 2014 9th International Conference on Industrial and Information Systems (ICIIS), IEEE, Gwalior, India, 15–17 December 2014; pp. 1–6. [Google Scholar]
- Kuremoto, T.; Kimura, S.; Kobayashi, K.; Obayashi, M. Time series forecasting using a deep belief network with restricted Boltzmann machines. Neurocomputing 2014, 137, 47–56. [Google Scholar] [CrossRef]
- Zhang, J.; Zheng, Y.; Sun, J.; Qi, D. Flow prediction in spatio-temporal networks based on multitask deep learning. IEEE Trans. Knowl. Data Eng. 2019, 32, 468–478. [Google Scholar] [CrossRef]
- Sun, S.; Wu, H.; Xiang, L. City-wide traffic flow forecasting using a deep convolutional neural network. Sensors 2020, 20, 421. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Fu, R.; Zhang, Z.; Li, L. Using LSTM and GRU neural network methods for traffic flow prediction. In Proceedings of the 2016 31st Youth Academic Annual Conference of Chinese Association of Automation (YAC), IEEE, Wuhan, China, 11–13 November 2016; pp. 324–328. [Google Scholar]
- Shi, R.; Du, L. Multi-Section Traffic Flow Prediction Based on MLR-LSTM Neural Network. Sensors 2022, 22, 7517. [Google Scholar] [CrossRef] [PubMed]
- Wei, W.; Wu, H.; Ma, H. An autoencoder and LSTM-based traffic flow prediction method. Sensors 2019, 19, 2946. [Google Scholar] [CrossRef] [Green Version]
- Wei, Y.; Liu, H. Convolutional Long-Short Term Memory Network with Multi-Head Attention Mechanism for Traffic Flow Prediction. Sensors 2022, 22, 7994. [Google Scholar] [CrossRef]
- Chang, L.; Chen, W.; Huang, J.; Bin, C.; Wang, W. Exploiting multi-attention network with contextual influence for point-of-interest recommendation. Appl. Intell. 2021, 51, 1904–1917. [Google Scholar] [CrossRef]
- Zhou, Q.; Chen, N.; Lin, S. FASTNN: A Deep Learning Approach for Traffic Flow Prediction Considering Spatiotemporal Features. Sensors 2022, 22, 6921. [Google Scholar] [CrossRef] [PubMed]
- Luo, D.; Zhao, D.; Ke, Q.; You, X.; Liu, L.; Zhang, D.; Ma, H.; Zuo, X. Fine-grained service-level passenger flow prediction for bus transit systems based on multitask deep learning. IEEE Trans. Intell. Transp. Syst. 2020, 22, 7184–7199. [Google Scholar] [CrossRef]
- Hou, Y.; Deng, Z.; Cui, H. Short-term traffic flow prediction with weather conditions: Based on deep learning algorithms and data fusion. Complexity 2021, 2021, 6662959. [Google Scholar] [CrossRef]
- Schuster, M.; Paliwal, K.K. Bidirectional recurrent neural networks. IEEE Trans. Signal Process. 1997, 45, 2673–2681. [Google Scholar] [CrossRef] [Green Version]
- Li, Y.; Yu, R.; Shahabi, C.; Liu, Y. Diffusion convolutional recurrent neural network: Data-driven traffic forecasting. arXiv 2017, arXiv:1707.01926. [Google Scholar]
- Hobeika, A.G.; Kim, C.K. Traffic-flow-prediction systems based on upstream traffic. In Proceedings of the VNIS’94-1994 Vehicle Navigation and Information Systems Conference, IEEE, Yokohama, Japan, 31 August–2 September 1994; pp. 345–350. [Google Scholar]
- Chen, T.; Guestrin, C. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM Sigkdd International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 13–17 August 2016; pp. 785–794. [Google Scholar]
- Ke, G.; Meng, Q.; Finley, T.; Wang, T.; Chen, W.; Ma, W.; Ye, Q.; Liu, T.Y. Lightgbm: A highly efficient gradient boosting decision tree. In Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, CA, USA, 4–9 December 2017; Volume 30. [Google Scholar]
- Cho, K.; Van Merriënboer, B.; Bahdanau, D.; Bengio, Y. On the properties of neural machine translation: Encoder-decoder approaches. arXiv 2014, arXiv:1409.1259. [Google Scholar]
- Hochreiter, S.; Schmidhuber, J. Long short-term memory. Neural Comput. 1997, 9, 1735–1780. [Google Scholar] [CrossRef] [PubMed]
- One-shot pruning of gated recurrent unit neural network by sensitivity for time-series prediction. Neurocomputing 2022, 512, 15–24. [CrossRef]
- Deng, J.; Chen, X.; Jiang, R.; Song, X.; Tsang, I.W. ST-Norm: Spatial and Temporal Normalization for Multi-Variate Time Series Forecasting. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD ’21, Singapore, 14–18 August 2021; Association for Computing Machinery: New York, NY, USA, 2021; pp. 269–278. [Google Scholar]
Research Gaps | Available Literature | Our Work |
---|---|---|
The impact of weather factors | Mostly rely on historical traffic volume changes as the basis for traffic flow prediction and overlook the influence of meteorological factors on traffic flow prediction in high-speed scenarios | Analyze the impact of meteorological factors on traffic flow prediction in high-speed scenarios and propose a multi-dimensional auxiliary information fusion model that captures meteorological factors for traffic flow prediction |
The influence of periodic patterns | Capture the periodicity in time dependence implicitly by modeling the sequence | Explicitly model the periodic changes in traffic flow, which can better capture the impact of periodic factors on traffic flow prediction |
Hangst | Metr-la | |
---|---|---|
City | Hangzhou | Los Angeles County |
Time span | February–October 2020 | March–June 2012 |
Time Interval | 5 min | 5 min |
Sensors | 200 | 207 |
Timestamp | Device Id | Property | Value |
---|---|---|---|
24 March 2021 17:31:53.158+08:00 | MD-TaiXing-85 | total_traffic | 20 |
24 March 2021 17:31:53.158+08:00 | MD-TaiXing-85 | average_speed | 83 |
24 March 2021 17:31:53.158+08:00 | WD-Mandrake-2 | visibility | 5000 |
24 March 2021 17:31:53.158+08:00 | WD-Mandrake-2 | precipitation | 0 |
24 March 2021 17:31:53.158+08:00 | WD-Mandrake-2 | nc_pavement_wet_coefficient | 65 |
Datasets | T | Metric | HA | SVR | XGBoost | LightGBM | RNN | LSTM | GRU | SCGRU | ST-Norm | MMLSTM |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Hangst | 30 min | RMSE | 26.80 | 61.36 | 61.64 | 61.33 | 44.82 | 54.31 | 48.27 | 22.27 | 21.46 | 19.98 |
MAE | 22.10 | 43.85 | 42.32 | 42.18 | 29.85 | 45.72 | 37.00 | 17.55 | 17.09 | 16.15 | ||
1 h | RMSE | 48.33 | 122.29 | 121.86 | 122.44 | 100.58 | 87.62 | 96.08 | 46.42 | 40.87 | 39.21 | |
MAE | 39.25 | 88.08 | 83.88 | 83.70 | 82.69 | 63.66 | 72.25 | 35.58 | 31.46 | 30.80 | ||
2 h | RMSE | 97.17 | 237.23 | 238.56 | 245.58 | 341.80 | 171.06 | 180.04 | 107.18 | 93.78 | 87.45 | |
MAE | 80.00 | 172.59 | 164.17 | 167.61 | 325.04 | 121.58 | 133.42 | 82.64 | 76.06 | 55.17 | ||
Metr-la | 30 min | RMSE | 102.12 | 109.29 | 106.89 | 106.62 | 89.95 | 84.00 | 84.64 | 76.30 | 56.60 | 45.20 |
MAE | 73.52 | 70.37 | 66.19 | 66.61 | 44.87 | 60.16 | 59.64 | 55.98 | 26.65 | 25.67 | ||
1 h | RMSE | 191.92 | 248.86 | 240.23 | 232.84 | 196.47 | 194.93 | 198.92 | 177.68 | 140.28 | 133.57 | |
MAE | 137.16 | 170.65 | 162.61 | 157.34 | 146.20 | 146.69 | 149.01 | 132.83 | 97.21 | 89.82 | ||
2 h | RMSE | 365.90 | 379.17 | 371.44 | 369.42 | 453.47 | 333.47 | 327.49 | 316.35 | 371.64 | 304.55 | |
MAE | 266.37 | 274.11 | 264.44 | 269.61 | 421.48 | 270.01 | 291.38 | 272.00 | 239.58 | 235.62 |
Dataset | Hangst | |||||
---|---|---|---|---|---|---|
T | 30 min | 1 h | 2 h | |||
Metric | MAE | RMSE | MAE | RMSE | MAE | RMSE |
Ours | 11.97 | 17.20 | 25.72 | 37.22 | 62.93 | 88.93 |
Ours-vh | 13.44 | 20.87 | 29.46 | 44.04 | 66.16 | 94.58 |
Ours-v | 13.78 | 20.98 | 27.82 | 41.80 | 63.62 | 94.10 |
Ours-h | 12.21 | 17.81 | 28.10 | 40.33 | 66.06 | 92.19 |
Dataset | Metr-la | |||||
T | 30 min | 1 h | 2 h | |||
Metric | MAE | RMSE | MAE | RMSE | MAE | RMSE |
Ours | 27.51 | 51.66 | 74.27 | 126.10 | 182.01 | 286.97 |
Ours-h | 28.39 | 51.71 | 75.86 | 125.92 | 193.37 | 301.02 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Li, B.; Xiong, J.; Wan, F.; Wang, C.; Wang, D. Incorporating Multivariate Auxiliary Information for Traffic Prediction on Highways. Sensors 2023, 23, 3631. https://doi.org/10.3390/s23073631
Li B, Xiong J, Wan F, Wang C, Wang D. Incorporating Multivariate Auxiliary Information for Traffic Prediction on Highways. Sensors. 2023; 23(7):3631. https://doi.org/10.3390/s23073631
Chicago/Turabian StyleLi, Bao, Jing Xiong, Feng Wan, Changhua Wang, and Dongjing Wang. 2023. "Incorporating Multivariate Auxiliary Information for Traffic Prediction on Highways" Sensors 23, no. 7: 3631. https://doi.org/10.3390/s23073631
APA StyleLi, B., Xiong, J., Wan, F., Wang, C., & Wang, D. (2023). Incorporating Multivariate Auxiliary Information for Traffic Prediction on Highways. Sensors, 23(7), 3631. https://doi.org/10.3390/s23073631