An Ant Colony Genetic Algorithm Based on Pheromone Diffusion

The conflict between exploring and exploiting is an unsolved problem for genetic algorithm. This paper proposed an ant colony genetic algorithm based on pheromone diffusion. The algorithm selects several sources from every sub-population, and these sources will be reserved to next population. Every source produces pheromones and diffuses them around, aim to guide the individual optimization. This strategy accelerates convergence speed with not damage its global search capability. From the experimental optimization results, the proposed method has better performance in global optimal solution precision and convergent speed.