New Hybrid Meta-heuristic to optimize the low exploitation of Discrete Adapted Dragonfly Algorithm for solving TSP
Atmane Ayoub Mansour Bahar, Kamel Soaïd Ferrahi, Samy Salma, Mohamed Yacine Touahria Miliani, Souhail Abdelmouaiz Sadat, Karim Laouchedi · HAL (Le Centre pour la Communication Scientifique Directe) · 2024
Optimization problems drive the use of efficient algorithms. Swarm intelligence algorithms, like the Dragonfly Algorithm (DA), have proven effective. However, the original DA is not adapted for discrete problems like the traveling salesman problem (TSP). Existing adaptations showed limited effectiveness, especially for large instances. To overcome this, a new hybrid meta-heuristic algorithm is proposed. It combines the Discrete Adapted DA with Simulated Annealing Search, resulting in significantly improved efficiency. The algorithm demonstrates promising results on large-scale TSP instances from the TSPLIB data set.