Article ID: | iaor1998329 |
Country: | United Kingdom |
Volume: | 48 |
Issue: | 3 |
Start Page Number: | 295 |
End Page Number: | 305 |
Publication Date: | Mar 1997 |
Journal: | Journal of the Operational Research Society |
Authors: | Hertz A., Costa D. |
Keywords: | graphs |
In the last few years researchers have shown how insect colonies can be seen as a natural model of collective problem solving. The analogy between the behaviour of ants looking for food and the well known travelling salesman problem has recently given rise to promising solution methods. This paper presents an evolutionary search procedure for tackling assignment type problems. The algorithm repeatedly constructs feasible solutions of the problem under study by taking account of two complementary notions, namely the trace factor and the desirability factor. The use of such search principles will be illustrated for graph colouring problems. The results obtained have proven satisfactory when compared with those existing in the literature.