Dataset and Implementation Codes for Resilient Real-Time Metro Rerouting Using Adaptive Bidirectional Heuristics and Graph Neural Networks
Description
This repository contains the dataset, source codes, trained graph models, and experimental outputs used in the research article entitled: "Resilient Real-Time Metro Rerouting with Adaptive Bidirectional Heuristics and Graph Neural Network-Based Edge Prediction" The provided materials support the implementation and evaluation of the proposed Adaptive Bidirectional Heuristic (ABH) routing framework integrated with Graph Neural Network (GNN)-based edge weight prediction. The repository includes metro network data, congestion information, routing algorithm implementations, GNN training codes, trained graph models, and experimental evaluation scripts. The provided files enable reproduction of the reported experiments and verification of the obtained results.
Files
Steps to reproduce
1. Install the required Python packages listed in requirements.txt. 2. Prepare the metro graph using the files provided in the DataSet directory. 3. Run GNN.py to construct the graph representation. 4. Run training-GNN-revised.py to train the GNN-based edge weight prediction model. 5. Execute FinalABH.py to perform routing using the predicted edge weights. 6. Run the evaluation scripts in the resultAndDiscussion directory to reproduce the reported experimental results.