BEST-WORST Ant System

The BEST-WORST Ant System (BWAS) algorithm discussed in this paper achieves a strong exploitation of the search history by allowing both the best solutions and the worst solutions to change pheromone during the pheromone trail update. It not only makes effective use of the positive feedback of iteration (global)-best ant, but also makes full use of the negative feedback of iteration (global)-worst ant. It improves the efficiency significantly. The use of a rather simple mechanism for limiting the strengths of the pheromone trails effectively avoids premature convergence of the search. Experimental results on TSP show that The BWAS algorithm has a better global searching ability, higher convergence speed and solution diversity than that of classical ACO algorithm.