Hybrid Grey Wolf and Genetic Algorithm for the Flow Shop Scheduling Problem

Mourad Mzili; Mouna Torki; Toufik Mzili; Maad M. Mijwil; Mohammed Benzakour Amine; Andres Annuk; Abderrahim Waga

doi:10.15157/IJITIS.2025.8.3.666-686

Authors

Mourad Mzili Faculty of Science, Chouaib Doukkali University, El Jadida, Morocco https://orcid.org/0009-0002-1985-480X
Mouna Torki Faculty of Science, Chouaib Doukkali University, El Jadida, Morocco https://orcid.org/0009-0002-1293-4014
Toufik Mzili Faculty of Science, Chouaib Doukkali University, El Jadida, Morocco https://orcid.org/0000-0002-5733-3119
Maad M. Mijwil College of Administration and Economics, Al-Iraqia University, Baghdad, Iraq, Computer Techniques Engineering Department, Baghdad College of Economic Sciences University, Baghdad, Iraq, Faculty of Engineering, Canadian Institute of Technology, Albania https://orcid.org/0000-0002-2884-2504
Mohammed Benzakour Amine Faculty of Science, Chouaib Doukkali University, El Jadida, Morocco https://orcid.org/0000-0003-1272-2441
Andres Annuk Institute of Forestry and Engineering, Estonian University of Life Sciences, Tartu, Estonia https://orcid.org/0000-0001-5464-0415
Abderrahim Waga Faculty of Sciences, Moulay Ismail University, Meknes, Morocco https://orcid.org/0000-0002-2672-3587

DOI:

https://doi.org/10.15157/IJITIS.2025.8.3.666-686

Keywords:

Flow Shop Scheduling, Grey Wolf Optimization, Genetic Algorithm, Hybrid Metaheuristics, Makespan Minimization, , Combinatorial Optimization, Manufacturing Systems

Abstract

The Flow Shop Scheduling Problem (FSSP), a pivotal NP-hard combinatorial optimization challenge, is central to enhancing manufacturing efficiency by minimizing makespan across n jobs and m machines. This study introduces a novel hybrid metaheuristic that integrates Grey Wolf Optimization (GWO) for robust global exploration with Genetic Algorithm (GA) for precise local exploitation, augmented by adaptive crossover, mutation, and 2-opt local search, addressing a significant gap in synthesizing swarm intelligence and evolutionary techniques for permutation-based scheduling. Evaluated on 13 Taillard benchmark instances (20-200 jobs, 5-20 machines) over 50 runs, the GWO-GA algorithm demonstrates superior performance compared to established metaheuristics, including SGA, HMSA, NEH, DDE-PFS, DSADE-PFS, and DSADEPFS, with statistical validation via ANOVA and Tukey HSD tests. The study highlights the algorithm's robust convergence and scalability, marking a key contribution to scheduling optimization. Its ability to outperform existing methods underscores its practical significance, while computational overhead for large instances suggests future exploration of parallelization and multi-objective enhancements.