GSO: An Improved PSO Based on Geese Flight Theory
Formation flight of swan geese is one type of swarm intelligence developed through evolution by natural selection. The research on its intrinsic mechanism has great impact on the bionics field. Based on previous research achievements, extensive observation and analysis on such phenomenon, five geese-flight rules and hypotheses are proposed in order to form a concise and simple geese-flight theory framework in this paper. Goose Swarm Optimization algorithm is derived based on the Standard Particle Swam Optimization algorithm. Experimental results show that GSO algorithm is superior in several aspects, such as convergence speed, convergence precision, robustness and etc. The theory offers the in-depth explanations for the performance superiority. Moreover, the rules and hypotheses for formation flight adhere to all five basic principles of swarm intelligence. Therefore, the proposed geese-flight theory is highly rational and has important theoretical innovations, and GSO algorithm can be utilized in a wide range of applications.
KeywordsGeese Theory Goose Swarm Optimization Particle Swarm Optimization Swarm Intelligence
Unable to display preview. Download preview PDF.
- 1.Kennedy, J., Eberhart, R.C.: Particle Swarm Optimization. In: Proceedings of the 1995 IEEE International Conference on Neural Networks, pp. 1942–1948. IEEE Press, Perth (1995)Google Scholar
- 3.Shi, Y.H., Eberhart, R.C.: Empirical Study of Particle Swarm Optimization. In: Proceeding of Congress on Evolutionary Computation, pp. 1945–1949. IEEE Press, Piscataway (1999)Google Scholar
- 4.Shi, Y.H., Eberhart, R.C.: A Modified Particle Swarm Optimizer. In: Proceedings of the 1998 IEEE International Conference on Evolutionary Computation, pp. 69–73. IEEE Press, Anchorage (1998)Google Scholar
- 5.Fukuyama, Y.: Fundamentals of Particle Swarm Techniques. IEEE Power Engineering Society 45–51 (2002)Google Scholar
- 9.Liu, J.Y., Guo, M.Z., Deng, C.: GeesePSO: An Efficient Improvement to Particle Swarm Optimization. Computer Science 33, 166–168 (2006)Google Scholar