Performance Evaluation of Emerging Meta‐Heuristic Algorithms on Vehicle Routing Problem

ABSTRACT This research provides a comprehensive evaluation of seven emergent meta‐heuristic algorithms, including flying fox optimization (FFO), Giza pyramids construction (GPC), Harris Hawks optimizer (HHO), red deer algorithm (RDA), whale optimization algorithm (WOA), mayfly optimization algorithm (MOA), and stochastic paint optimizer (SPO) applied to the vehicle routing problem (VRP). The algorithms were implemented in MATLAB and assessed based on solution quality, execution time, and convergence rate across small, medium, and large‐scale problems. The evaluation revealed significant performance variations among these algorithms. WOA consistently achieved top ranks in small and medium‐scale problems, demonstrating its robustness and efficiency. In contrast, GPC excelled in large‐scale problems, outperforming other algorithms in handling complex and extensive datasets. SPO, however, consistently ranked lowest across all scales, indicating its limited effectiveness for VRP under the tested conditions. The study employed the Shannon Entropy method for weighting the evaluation criteria and a multi‐criteria decision‐making method for the final ranking of the algorithms, providing a structured and comprehensive assessment approach. The findings suggest that WOA is the most effective algorithm, offering reliable and high‐quality solutions with efficient execution times and convergence rates, while SPO requires significant enhancements. These insights are valuable for practitioners and managers in logistics and supply chain management, guiding the selection of appropriate algorithms based on problem scale. The research also opens avenues for future work, including the refinement of lower‐performing algorithms, comprehensive testing with broader datasets, advanced parameter optimization, and exploration of algorithm applicability in other domains, such as scheduling and resource allocation. This study not only benchmarks the performance of emerging meta‐heuristic algorithms on VRP but also lays a foundation for future advancements in optimization techniques.