Verifying Usefulness of Ant Colony Community for Solving Dynamic TSP
The paper describes Ant Colony Communities (ACC) and verifies their usefulness for Dynamic Travelling Salesman Problem (DTSP). DTSP is a version of the classical TSP in which the distance matrix change in time. The ACC consists of a set of separate ant colonies with a server that coordinates their work and sends them cargos of data for processing. The colonies could be distributed over many computers working in a LAN or even over Internet. Such a mode of operation is especially useful for dynamic tasks where solutions must catch up with the changing environment. The ACC is used for the regular ACO and its version designed for dynamic environments: PACO and Immigrant ant colonies. The experiments show that for all types Ant Colonies the introduction of the community boots the performance. The routes are far shorter than in the case of original colonies.
KeywordsDynamic TSP Ant Colony Community PACO Immigrant based colonies Parallel implementation of ACO
- 1.Antosiewicz, M., Koloch, G., Kamiński, B.: Choice of best possible metaheuristic algorithm for the travelling salesman problem with limited computational time: quality, uncertainty and speed. J. Theor. Appl. Comput. Sci. 7(1), 46–55 (2013)Google Scholar
- 2.Dorigo, M., Stuetzle, T.: Ant colony optimization: overview and recent advances, IRIDIA - Technical Report Series, Technical report No. TR/IRIDIA/2009-013 (2009)Google Scholar
- 4.Dorigo, M.: Optimization, Learning and Natural Algorithms, Ph.D. thesis, Politecnico di Mila-no, Italie (1992)Google Scholar
- 13.Psarafits, H.N.: Dynamic vehicle routing: Status and Prospects. National Technical Annals of Operations Research, University of Athens, Greece (1995)Google Scholar
- 15.Chirico, U.: A Java framework for ant colony systems. In: Ants2004: Forth International Workshop on Ant Colony Optimization and Swarm Intelligence, Brussels (2004)Google Scholar