Hybrid Flexible Flow-shop Optimization Thesis

This repository contains all algorithms, implementations, experiments, and analysis from my thesis, which aims to compare multiple metaheuristics and approaches to the Hybrid Flexible Flow-shop (HFFS) problem, a popular production scheduling problem. For a comprehensive overview of the research, methodology, and results, please refer to the thesis document.

Running Experiments

• Genetic Algorithm Experiments:
  • Execute src/ga_experiments.py to run new experiments for the genetic algorithm.
• ALNS Experiments:
  • Execute src/alns_experiments.py to run new experiments for the ALNS (Adaptive Large Neighborhood Search) algorithm.
• Both scripts utilize problem instances located in src/input/.

• Make sure you have first run pip install -r requirements.txt to install all needed dependencies

Analysis and Replication

• The analysis/ directory contains complete analysis notebooks for the results.
• To replicate the results of the thesis, execute the following notebooks:
  • analysis/alns_analysis.ipynb: analysis of the ALNS execution results
  • analysis/ga_analysis.ipynb: analysis of the GA execution results
  • analysis/optuna_analysis.ipynb: analysis of the parameter tuning results of the GA

Instance Generation

• The instance_generation/ directory holds all scripts used for the generation and sampling of problem instances.

---

For further details, please consult the thesis document included in this repository.