Dataset for Benchmarking the Swallow-Bat Algorithm on a Threat-Aware Container Truck Routing Problem
Description
This dataset provides comprehensive benchmark instances for threat-aware container truck routing problems. It includes three modified VRP benchmark instances (21, 33, and 51 customers) enhanced with static spatial threat zones, plus a real-world East African Community transport corridors case study with 25 cities and 16 evidence-based threat zones. The dataset contains complete problem specifications, threat zone definitions, travel cost matrices from OSRM/OpenStreetMap, and performance metrics for six metaheuristic algorithms across 30 independent runs. Data formats include CSV files for node data, threat zones, and travel matrices, plus text files for problem parameters.
Files
Steps to reproduce
1. Download the complete dataset and extract it to local directory 2. Install Python 3.8+ with numpy, pandas, matplotlib, scipy 3. Run: python source_code/main_experiment.py --instance all --runs 30 4. Compare results with the provided experimental_results/ files 5. Verify SBA achieves lowest mean cost and <2% Coefficient of Variation 6. Full reproduction requires 8-12 hours of computation time
Institutions
- Central South University Railway Campus
Categories
Funders
- Natural Science Foundation of Hunan ProvinceGrant ID: 2023JJ30710-2022JJ31020