Improving performance of ACO algorithms using crossover mechanism based on mean of pheromone tables

Ant Colony Optimization (ACO) Algorithms have been used to solve many optimization problems in various fields and several algorithms have been proposed based on ACO metaheuristic in the literature. This paper proposes a simple crossover mechanism based on mean of pheromone tables for ACO algorithms. Main purpose of the crossover operation is to produce solutions or individuals having greater performance than their parents by selecting useful parts. Original ACO Algorithms don't have crossover. Method that we developed employs more than one ant colonies and also solutions. Suitable low-cost average based operations are then applied to pheromone tables obtained after several iterations as crossover operator. Algorithm is tested on Traveling Salesman Problem using some benchmark problems from TSPLIB and results are presented. Our experiments and comparisons show that crossover mechanism improves the performance of ACO Algorithms.