Akademik Veri Yönetim Sistemi
Cantürk D., Sermimer C.(Yürütücü)
Yükseköğretim Kurumları Destekli Proje, Üniversite Destekli Diğer Projeler, 2026 - 2026
The Quadratic Assignment Problem (QAP) remains one of the most challenging combinatorial optimization problems due to its NP-hard nature and extensive applications in facility layout, circuit design, and logistics. While metaheuristic algorithms such as Harris Hawks Optimization (HHO) and Genetic Algorithms (GA) have shown promise for solving large-scale QAPs, they face significant computational challenges related to premature convergence and excessive runtime. This paper presents a comparative study of two enhanced approaches: a GPU-accelerated HHO algorithm with parallelized tabu search initialization, and a diversity-controlled GA utilizing novel injection mechanisms. The GPU implementation leverages OpenCL for parallel population generation, reducing computational overhead while maintaining solution quality. The diversity-controlled GA employs systematic population injection strategies to prevent premature convergence and maintain exploration capabilities throughout the search process.