Loading...

Predicting Origin-Destination Matrices in Expanding Large-Scale Urban Networks Using Graph Neural Networks

Parsi, Amir Hossein | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57528 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Amini, Zahra
  7. Abstract:
  8. With the increase in population and the expansion of cities, the management of transportation infrastructure has become highly significant. Accurate prediction of origin-destination matrices, which indicate the number of trips between different areas, plays a crucial role in urban planning, traffic management, and infrastructure development. The aim of this research is to design and implement a model based on a graph neural network to estimate origin-destination matrices in large and expanding networks. In this study, the features of origin and destination nodes are extracted and combined using two separate graph neural networks, and then the combined features for each origin-destination pair are input into a neural network to predict the number of trips. Finally, after validating the model with real data, an efficient and accurate model for various situations is proposed. The output of this model is compared and evaluated against the gravity model and machine learning-based models, highlighting its advantages. To assess the model's performance, data from the years 2011 and 2016 for the Greater Toronto and Hamilton area was used. The results show that the proposed model, with an R-squared of 93% in predicting the 2016 origin-destination matrix, outperforms other models. Additionally, the R-squared of this model in predicting the number of trips to and from new areas in 2016 is around 90%, which is a significant improvement compared to the 78% R-squared of other models
  9. Keywords:
  10. Origin-Destination Matrix ; Graph Neural Network ; Machine Learning ; Network Expansion ; Transportation Planning ; Urban Transportation Network

 Digital Object List

 Bookmark

No TOC