Effectiveness of an ACO Algorithm with Levy Flight for Large-Scale Constraint Satisfaction Problems

To solve large-scale and hard constraint satisfaction problems (CSPs), ant colony optimization (ACO) has been studied as one of the metaheuristics. Lévy ACO, which incorporates Lévy Flight (LF) into ACO, has been proposed as a more efficient method for solving CSPs. However, Lévy ACO has a fixed parameter that determines the frequency of LF use (LF parameter), throughout the search process, leading to reduce the efficiency of the search. In this paper, we propose a method to dynamically adjust the LF parameter according to the search progresses. We demonstrate that our method can solve large-scale and hard graph coloring problems, which is one of CSPs, more efficiently than the previous methods.