Parallel ACO algorithms for 2D Strip Packing

In this paper we present a study of a parallel Ant Colony System (ACS) for the two-dimensional strip packing problem. In our computational study, we emphasize the in uence of the incorporation of the received information in the target subcolony. Colonies send their best solutions instead of sending information from the matrix of pheromones, as happens in traditional parallel ACS. The solution arriving to a colony can provide further exploitation around promising solutions as this arrived solution can be used in both, the local update of the pheromone trail and the construction solution process of an ant. The aim of the paper is to report experimental results on the behavior of different types of parallel ACS algorithms, regarding solution qualities and parallel performance. Presentado en XI Workshop Agentes y Sistemas Inteligentes (WASI) Red de Universidades con Carreras en Informática (RedUNCI)