Abstract:

In solving small- to medium-scale travelling salesmanproblems (TSPs) of both symmetric and asymmetric types, thetraditional ant colony optimization (ACO) algorithm could workwell, providing high accuracy and satisfactory efficiency. However,when the scale of the TSP increases, ACO, a heuristic algorithm,is greatly challenged with respect to accuracy and efficiency. Anovel pheromone-trail updating strategy that moderately reducesthe iteration time required in real optimization problem-solvingis proposed. In comparison with the traditional strategy of theACO in several experiments, the proposed strategy shows advantagesin performance. Therefore, this strategy of pheromone-trailupdating is proposed as a valuable approach that reduces thetime-complexity and increases its efficiency with less iteration timein real optimization applications. Moreover, this strategy is especiallyapplicable in solving the moderate large-scale TSPs basedon ACO.