Traffic Flow Prediction based on Bi LSTM and Attention

Wei Dong ； Wei Sun

《International Core Journal of Engineering》 8卷3期 (2022/03) Pp. 439-444

https://doi.org/10.6919/ICJE.202203_8(3).0057

摘要

In order to improve the accuracy of traffic time prediction, a combined model based on bilstm and attention is established. That is to add attention mechanism to the bilstm model, so that the model can give different weights to different time steps. Experiments show that the bilstm-att model with attention mechanism has better prediction ability than the bilstm model.

關鍵字

Traffic Forecast ； Spatiotemporal Correlation ； Self Attention Mechanism ； Graph Convolution Network

參考文獻

M. Lippi, M. Bertini, and P. Frasconi, “Short-term traffic flow forecasting: An experimental comparison of time-series analysis and supervised learning,” IEEE Transactions on Intelligent Transportation Systems, vol. 14, no. 2, pp. 871–882, 2013.

I. Okutani and Y. J. Stephanedes, “Dynamic prediction of traffic volume through kalman filtering theory,” Transportation Research Part B, vol. 18, no. 1, pp. 1–11, 1984.

B. M. Williams and L. A. Hoel, “Modeling and forecasting vehicular traffic flow as a seasonal arima process: Theoret- ical basis and empirical results,” Journal of Transportation Engineering, vol. 129, no. 6, pp. 664–672, 2003.

Y. Gao and S. Sun, “Multi-link traffic flow forecasting using neural networks,” in International Conference on Natural Computation, 2010, pp. 398–401.

Y. Lv, Y. Duan, W. Kang, Z. Li, and F. Y. Wang, “Traffic flow prediction with big data: A deep learning approach,” IEEE Transactions on Intelligent Transportation Systems, vol. 16, no. 2, pp. 865–873, 2015.