Worm optimization for the traveling salesman problem

Abstract

In this research, a new metaheuristic called Worm Optimization (WO) is proposed, based on the foraging behaviors of Caenorhabditis elegans (Worms). In particular, the algorithm will mimic the behaviors of worms including finding food, avoiding toxins, interchanging between solitary and social foraging styles, alternating between food exploiting and seeking, and entering a stasis stage. WO effectiveness is illustrated on the traveling salesman problem (TSP), a known NP- hard problem, and compared to well-known naturally inspired algorithms using existing TSP data. The computational results reflected the superiority of WO in all tested problems. Furthermore, this superiority improved as problem sizes increased, and WO attained the global optimal solution in all tested problems within a reasonable computational time.