Understanding the predictability of path flow distribution in urban road networks using an information entropy approach
Lists
Only you see this
Abstract
Predicting the distributions of path flow between origin-destination (OD) pairs in an urban road network is crucial for developing efficient traffic control and management strategies. Here, we use the large-scale taxi GPS trajectory data of San Francisco and Shenzhen to study the predictability of path flow distribution in urban road networks. We develop an approach to project the time-varying path flow distributions into a high-dimensional space. In the high-dimensional space, information entropy is used to measure the predictability of path flow distribution. We find that the distributions of path flow between OD pairs are in general characterized with a high predictability. In addition, we analyze the factors affecting the predictability of path flow distribution. Finally, an n-gram model incorporating high-order gram and low-order gram is proposed to predict the distribution of path flow. A relatively high prediction accuracy is achieved.
BibTeX
@inproceedings{guo2024understanding,
author = {Bao Guo and Zhiren Huang and Zhihao Zheng and Fan Zhang and Pu Wang},
title = {{Understanding the predictability of path flow distribution in urban road networks using an information entropy approach}},
booktitle = {Multimodal},
year = {2024},
doi = {10.1016/j.multra.2024.100135},
}Loading related content...