A Novel Two-Level Hybrid Algorithm for Multiple Traveling Salesman Problems
Multiple traveling salesmen problem is a NP-hard problem. The method for solving the problem must arrange with reason all cities among traveling salesman and find optimal solution for every traveling salesman. In this paper, two-level hybrid algorithm is put forward to take into account these two aspects. Top level is new designed genetic algorithm to implement city exchange among traveling salesmen with the result clustered by k-means. Bottom level employs branch-and-cut and Lin-kernighan algorithms to solve exactly sub-problems for every traveling salesman. This work has both the global optimization ability from genetic algorithm and the local optimization ability from branch-and-cut.
Keywordsmultiple traveling salesmen problem two-level hybrid algorithm crossover operator chromosome encoding
Unable to display preview. Download preview PDF.
- 3.Greco, F.: Traveling Salesman Problem. InTech (2008)Google Scholar
- 5.Diaby, M.: Linear Programming Formulation of the Multi-Depot Multiple Traveling Salesman Problem with Differentiated Travel Costs. In: Davendra, D. (ed.) Traveling Salesman Problem, Theory and Applications. InTech (2010)Google Scholar
- 7.Lin, D.-M., Wang, D., Li, Y.: Two-level degradation hybrid algorithm for multiple traveling salesman problem. Application Research of Computers 28(8), 2876–2879 (2010)Google Scholar
- 10.Schabauer, H., Schikuta, E., Weishaupl, T.: Solving Very Large Travelling Salesman Problems by SOM parallelization on cluster architecture. In: Proceedings of the Sixth International Conference on Parallel and Distributed Computing Applications and Technologies (PDCAT), pp. 954–958 (2005)Google Scholar
- 13.Traveling Salesman Problems Library, http://www.iwr.uni-heidelberg.de/groups/comopt/software/TSPLIB95/