Pheromone landscapes in numerical optimization

In this paper we present a new algorithm based on ant colony metaphor for examining numerical optimization problems. The ants’ search behavior consists of a parallel searches on distributed randomly points. Using synergetic effect and tandem running ants try to calibrate with another ants to exchange information about the fruitful positions of search space. The position of the searching point in each region is periodically moved to more fruitful regions according to the value of pheromone and agents try to analyse the neighborhood of this point. This algorithm is tested on several testable functions and finally we conclude the achieved results.