A new optimization algorithm based on ant colony system with density control strategy

Abstract

A new optimization algorithm based on the ant colony system is presented by adopting the density control strategy to guarantee the performance of the algorithm. In each iteration of the algorithm, the solutions are selected to have mutation operations according to the quality and distribution of the solution. Experimental results on the traveling salesman problem show that our algorithm can not only get diversified solutions and higher convergence speed than the Neural Network Model and traditional ant colony algorithm, but also avoid the stagnation and premature problem. © Springer-Verlag Berlin Heidelberg 2006.