Performance Evaluation of a Modified Version of the White Shark Optimizer for Solving the Traveling Salesman Problem

The Traveling Salesman Problem (TSP) is a long-standing and widely recognized issue in combinatorial optimization. It is known for presenting considerable computational obstacles due to its classification as an NP-hard problem. The crux of the TSP is to find the most efficient, i.e., shortest, route a traveling salesperson can adopt to visit a pre-determined list of cities, returning to the point of origin without revisiting any cities. In this context, our paper proposes a novel and innovative approach to address the TSP, employing the White Shark Optimizer (WSO) algorithm as the core methodology. The WSO is a relatively recent addition to the class of swarm intelligence algorithms. The design and operation of the WSO draw inspiration from the behavioral patterns observed in the hunting strategies employed by white sharks. This concept separates it from other traditional methods. We implemented several adaptations to tailor the WSO to the specific requirements of the TSP. These include viewing the cities in the TSP as individual particles. Furthermore, to integrate the unique wavy movement patterns of the sharks into the context of the TSP, we implemented a modified version of the velocity update function. This newly developed methodology was tested rigorously on benchmark TSP instances to validate its efficacy and performance. The findings from our experiments indicate that our unique approach not only holds its ground against more traditional algorithms but also provides competitive, if not superior, solutions. This suggests it has considerable potential to tackle intricate and complex optimization problems.

کلیدواژه ها:

Traveling Salesman Problem (TSP) ، White Shark Optimizer (WSO) ، Particle Swarm Optimization (PSO) ، Metaheuristic Algorithms (MAs) ، Swarm Intelligence (SI) ، Optimization Problems ، Combinatorial Optimization