Multi-Thread Memory Particle Swarm Optimization for Dynamic Optimization Problems

Abstract:

This paper proposes a multi-thread-memory particle swarm optimization (MTM-PSO) for dynamic optimization problems (DOPs). It introduces multi-thread memory to multi-swarm approaches to deal with DOPs. External memory and multi-swarm approaches are adopted. The best particles near the peaks in each environment are saved in the memory according to storage strategy of update and implement. Multi-thread memory makes full use of local and global optima found in each environment. It makes the good information in the previous evolution transfer to the current population adequately. Using the multi-thread memory and its storage strategy, the particles in the memory are always the best information of local and global optima found until in the current envionment. The information will benefit the future evolution. Experiments on the moving peak benchmark (MPB) for different numbers of peaks testiy MTM-PSO has better performance than dynamic optimization algorithms only with multi-swarm approach.

• Periodical: Applied Mechanics and Materials (Volumes 571-572)
• Pages: 232-236
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.571-572.232
• Online since: June 2014
• Authors: Li Chen*, Ju Ming Liu, Jun Mei Ouyang
• Keywords: Dynamic optimization, Multiswarm Particle Swarm Optimization, Multi-Thread Memory

References

[1] Branke J. Evolutionary Optimization in Dynamic Environments. Boston: Kluwer Academic Publishers (2002).

[2] Goldberg D E, Smith R E. Nonstationary function optimization using genetic algorithms with dominance and diploidy. Proceedings of the Second International Conference on Genetic Algorithms on Genetic algorithms and their application(1987), p.59.

DOI: 10.1049/cp:19971212

[3] Branke J. Memory enhanced evolutionary algorithms for changing optimization problems. Proc. 1999 Congr. Evol. Comput(1999), p.1875.

[4] Yang S. Associative memory scheme for genetic algorithms in dynamic environments. Proc. of the EvoWorkshops 2006(2006), p.788.

[5] Parrott D, Li X. IEEE Trans. Evol. Comput. Vol. 10 (2006), p.440.

[6] Yang S, Li C. IEEE Transactions on Evolutionary Computation. Vol. 14 (2010), p.959.

[7] Blackwell T M, Branke J. IEEE Transactions on Evolutionary Computation. Vol. 10(2006), p.459.

[8] Bird S, Li X. Using regression to improve local convergence. Proceddings of Congress on Evolutionary Computation (CEC'07). (2007), p.592.

DOI: 10.1109/cec.2007.4424524

[9] Blackwell T M, Branke J. IEEE Transactions on Evolutionary Computation. Vol. 10 (2006), p.459.