Influence of parameters of the ant colony algorithm on the traveling salesman problem solution

Transportation of many freights can be given as the travelling salesman problem, when freight is delivered from one distribution center to customers during one trip. There are used exact, heuristic and metaheuristic methods for the solving this problem. It was chosen the ant colony algorithm from metaheuristic methods because it is close to the statement of the travelling salesman problem at the expense of its physical resemblance. However, the control parameters of the algorithm influent on its efficiency. Therefore, investigation of the searching of control parameter values, by which the algorithm will look for the optimal route as soon as possible, was carried out.The ant colony algorithm for the travelling salesman problem was implemented in MATLAB environment. At the first step for the networks with 15, 20, 25 and 30 nodes, that answer to the real delivery systems, there was determined the set of values of the control parameters, which ensure the largest efficiency of the ant colony algorithm –α=1,β=5,ρ=0,2. At the second step, there was determined minimal amount of iterations needed for searching of the optimal route for these networks. At the third step there was determined that insertion of three elite ants enabled to decrease amount of iterations for the optimal route searching (for example, for 20-nodes network from 307 to 69).Obtained results can be used not only for the travelling salesman problem solving but also for vehicle routing of small-batch trucking, when the ant colony algorithm considers additional conditions.