Short-Term Traffic Flow Forecasting Considering Upstream Traffic Information

DOI

10.2991/mecae-18.2018.86How to use a DOI?

Keywords

short-term traffic flow forecasting, K-NN model, upstream information.

Abstract

Timely and accurately short-term traffic flow forecasting is very important in the application of the Intelligent Transportation System (ITS). In this paper, an improved k-nearest neighbor (K-NN) model considering upstream traffic information was proposed to forecast short-term traffic flow. In this study, the traffic state of a road segment was described as a state matrix with upstream information instead of only a time series vector. And the weighted Euclidean distance gave different weight to target and upstream road segment was used to measure the similarity between the target state matrix and historical state matrix. The K-NN model was trained by the training data to determine the optimal K value. This study used the reverse distance weighted average method based on similarity of the neighbors to generate the forecasting traffic flow in the future time steps. The same traffic data was used to compare the improved model with three models.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Fei Kou
AU  - Weixiang Xu
AU  - Huiting Yang
PY  - 2018/03
DA  - 2018/03
TI  - Short-Term Traffic Flow Forecasting Considering Upstream Traffic Information
BT  - Proceedings of the 2018 International Conference on Mechanical, Electronic, Control and Automation Engineering (MECAE 2018)
PB  - Atlantis Press
SP  - 560
EP  - 564
SN  - 2352-5401
UR  - https://doi.org/10.2991/mecae-18.2018.86
DO  - 10.2991/mecae-18.2018.86
ID  - Kou2018/03
ER  -