臺灣博碩士論文加值系統

準確的交通預測可以幫助人們避免可能的交通擁堵。本文將使用交通部高速公路局資料庫，利用長短期記憶神經網路(Long Short Time Memory Neural Network, LSTM)，預測連續假日中國道 5 號的車速情況。從實驗結果可看出 LSTM 能提供良好的車速預測，可用於分析國道 5 號的交通狀況。

A precise forecast of traffic may help people avoiding possible traffic congestions. In this paper, we will apply the LSTM to predict the speed of vehicle flow in Freeway No. 5 by using the data from Freeway Bureau, MOTC. The results of experiments show that LSTM is able to supply a good prediction, and we may use it to analyze the traffic of Freeway No. 5.

1. 緒論……………………………………………………………………………………………………………1
1.1 研究動機與目的…………………………………………………………………………………1
1.2 論文架構………………………………………………………………………………………………2
2. 長短期記憶神經網路…………………………………………………………….……..2
2.1 時間序列…………………………………………………………………………..………………2
2.2 類神經網路…………………………………………………………………………………………3
2.3 循環神經網路…………………………………………………………………..….………3
2.4 長短期記憶神經網路……………………………………………………………………4
3. 實驗方法與步驟………………………………………………………………………………..7
3.1 數據說明……………………………………………………………………………………..…7
3.2 模型設計………………………………………………………………………………..………8
4. 實驗結果與分析…………………………………………..……………………………….11
4.1 連續假日— 4 天………………………………………………………………………..11
4.2 連續假日— 3 天………………………………………………………………….…..…16
4.3 連續假日— 2 天……………………………………………………………………………20
5. 結論……………………………………………………………………………………………………….…22
參考文獻………………………………………………………………………………………….……………23
附錄一: 連續假期 3 天中第 1 天的圖形……………….…………………24
附錄二: 連續假期 3 天中第 2 天的圖形…………….…….……………28
附錄三: 連續假期 2 天中第 1 天的圖形…………………….……………33
附錄四: 連續假期 2 天中第 2 天的圖形…………….…….……………36

[1] Ahmed , Cook, Analysis of freeway traffic time-series data by using Box-
Jenkins techniques, Transp. Res. Rec no. 722, pp. 1-9, 1979
[2] Wu, C.-H., Ho, J.-M., Lee, D.T., Travel-time prediction with support vector
regression, IEEE Transactions on Intelligent Transportation Systems 5 (4),
276–281, 2004
[3] Sepp Hochreiter, Jurgen Schmidhuber, LONG SHORT-TERM MEMORY, Neural
Computation 9(8):1735-1780, 1997
[4] YiFei Li, Han Cao, Prediction for tourism flow based on LSTM Neural Net-
-work, International Conference on Identification, Information and Knowled-
-ge in the Internet of Things, 2017
[5] Oinkina, http://colah.github.io/posts/2015-08-Understanding-LSTMs/,2015
[6] 交通部高速公路局交通資料庫，http://tisvcloud.freeway.gov.tw/