Dynamic Adjustment of Control Parameters in ACO for Constraint Satisfaction Problems

Abstract

Ant colony optimization (ACO) has been effective meta-heuristics to solve hard instances of constraint satisfaction problems (CSPs). In ACO based algorithms, it is generally necessary and difficult to set at least a pair of control parameters in advance. In this paper, we propose an ACO based algorithm that can dynamically adjust the control parameters during the search progress by adopting a particle swarm intelligence based approach. We demonstrate that dynamically adjustment of the parameters in our proposed method can be effective and efficiently solve large-scale and hard graph coloring problems which are one of typical examples of CSPs.