A Parallel Framework for Multi-Population Cultural Algorithm and Its Applications in TSP
This paper presents a novel parallel framework based on the Multi-Population Cultural Algorithm (MPCA) scheme for optimization problems. Contrary to the existing variants of Cultural Algorithm (CA), the proposed parallel framework for MPCA (PFMPCA) allows the use of any implemented metaheuristic both in a belief, and in a population space. Furthermore, the proposed approach permits CA to evolve simultaneously multiple population and belief sub-spaces, leveraging the dual inheritance mechanism and utilizing multi-population approach. Moreover, each sub-population (in population or belief space) is able to communicate between each other. PFMPCA has been implemented on Graphics Processing Units (GPUs) using CUDA programming model. The performance of the developed framework was evaluated using asymmetric Travelling Salesman Problem (ATSP). The MPCA for TSP implemented by means of the parallel framework proves to have an extensible architecture designed to accommodate changes and good performances.
KeywordsCultural Algorithm Multi-Population GPU computing CUDA architecture Travelling Salesman Problem Ant Colony Optimization Genetic Algorithm
The work was supported by statutory grant of the Wroclaw University of Science and Technology, Poland.
OU initiated and designed the study, supervised the work, made statistical tests. RT implemented the framework, performed the experiments. Both authors wrote and approved the final manuscript.
- 1.Ali, M.Z.: Using cultural algorithms to solve optimization problems with a social fabric approach. Ph.D. thesis, Wayne State University (2008)Google Scholar
- 4.Dong, J., Yuan, B.: GPU-accelerated standard and multi-population cultural algorithms. In: 2013 International Conference on Service Sciences (ICSS), pp. 129–133. IEEE (2013)Google Scholar
- 5.Dorigo, M.: Optimization, learning and natural algorithms. Ph.D. thesis, Politecnico di Milano, Italy (1992)Google Scholar
- 8.Guo, Y.N., Liu, D.: Multi-population cooperative particle swarm cultural algorithms. In: 2011 Seventh International Conference on Natural Computation (ICNC), vol. 3, pp. 1351–1355. IEEE (2011)Google Scholar
- 9.Hlynka, A.W., Kobti, Z.: Knowledge sharing through agent migration with multi-population cultural algorithm. In: FLAIRS Conference (2013)Google Scholar
- 10.Hlynka, A.W., Kobti, Z.: Heritage-dynamic cultural algorithm for multi-population solutions. In: 2016 IEEE Congress on Evolutionary Computation (CEC), pp. 4398–4404. IEEE (2016)Google Scholar
- 12.Kobti, Z., et al.: Heterogeneous multi-population cultural algorithm. In: 2013 IEEE Congress on Evolutionary Computation (CEC), pp. 292–299. IEEE (2013)Google Scholar
- 13.Nvidia: Nvidia CUDA (2017). http://nvidia.com/cuda
- 15.Reinhelt, G.: TSPLIB: a library of sample instances for the TSP (and related problems) from various sources and of various types (2017). http://comopt.ifi.uniheidelberg.de/software/TSPLIB95
- 16.Reynolds, R.G.: An introduction to cultural algorithms. In: Proceedings of the Third Annual Conference on Evolutionary Programming, Singapore, pp. 131–139 (1994)Google Scholar
- 17.Stützle, T., Hoos, H.: MAX-MIN ant system and local search for the traveling salesman problem. In: IEEE International Conference on Evolutionary Computation, pp. 309–314. IEEE (1997)Google Scholar
- 19.Xiao, S., Feng, W.: Inter-block GPU communication via fast barrier synchronization. In: 2010 IEEE International Symposium on Parallel & Distributed Processing (IPDPS), pp. 1–12. IEEE (2010)Google Scholar