A Particle Swarm-based Metaheuristic to Solve the Travelling Salesman Problem.
Alberto Gómez Gómez, Ricardo Mendoza-González, José Parreño Fernández, Raúl Pino Díez · 2006
This paper describes a new metaheuristic technique to solve the travelling salesman problem based on particle swarm using guided local search. The main contribution of the paper is that is develops a particle swarm-based strategy to solve combinatorial optimization problems in general, and the TSP in particular. It first establishes the steps to be followed in the new method and then goes on to test the results obtained on a range of symmetrical problems set by different authors in the field. The method’s results are then compared with results from a number of well known non-guided local search techniques, hoighlighting the fact that it is an enhancement on earlier techniques. Finally, it studies the contribution of particle swarm to the local search process search, where there is also an improvement over results from a simple local search. Experimental results suggest that the new method provided good results for symmetric TSP problems.