Predicted modified PSO with time-varying accelerator coefficients
by Xingjuan Cai, Yan Cui, Ying Tan
International Journal of Bio-Inspired Computation (IJBIC), Vol. 1, No. 1/2, 2009

Abstract: Cognitive and social learning factors are two important parameters associated with the performance of particle swarm optimisation significantly. Up to date, many selection strategies have been proposed aiming to improve either the performance or the population diversity. One of the most widely used improvements is the linear selection manner proposed by Ratnaweera in 2004. However, due to the complex nature of the optimisation problems, linear automation strategy may not work well in many cases. Since the large cognitive coefficient provides a large local search capability, whereas the small one employs a large global search capability, a new variant – predicted modified particle swarm optimisation with time-varying accelerator coefficients, in which the social and cognitive learning factors are adjusted according to a predefined predicted velocity index. If the average velocity of one particle is superior to the index, its social and cognitive parameters will chose a convergent setting, and vice versa. Simulation results show the proposed variant is more effective and efficient than other three variants of particle swarm optimisation when solving multi-modal high-dimensional numerical problems.

Online publication date: Mon, 26-Jan-2009

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Bio-Inspired Computation (IJBIC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com