Acceleration Based Particle Swarm Optimization for Graph Coloring Problem

Abstract The graph coloring problem is one of the combinatorial optimization problems. Although many heuristics and metaheuristics algorithm were developed to solve graph coloring problem but they have some limitations in one way or another. In case of tabu search, the algorithm becomes slow, if the tabu list is big. This is because lots of memory to keep the list and also a lot of time to travel through the list, is needed in each step of the algorithm. Simulated annealing has a big handicap when applied to graph coloring problem because there are lots of neighboring states that have the same energy value. The problem with ant colony optimization is that the number of ants that must be checked is n times bigger than other algorithms. Therefore, there will be a need of a large amount of memory and the computational time of this algorithm can be very large. A swarm intelligence based technique called as particle swarm optimization is therefore employed to solve the graph coloring problem. Particle swarm optimization is simple and powerful technique but its main drawback is its ability of being trapped in the local optimum. Therefore, to overcome this, an efficient Acceleration based Particle Swarm Optimization (APSO) is introduced in this paper. Empirical study of the proposed APSO algorithm is performed on the second DIMACS challenge benchmarks. The APSO results are compared with the standard PSO algorithm and experimental results validates the superiority of the proposed APSO.