Fractional Order Darwinian Particle Swarm Optimization by Couceiro Micael & Ghamisi Pedram

Author:Couceiro, Micael & Ghamisi, Pedram
Language: eng
Format: epub
Publisher: Springer International Publishing
Published: 2015-06-15T16:00:00+00:00

As a conclusion, FODPSO-based image segmentation is able to reach a slightly better fitness solution in less CPU processing time than its alternatives (DPSO and PSO). This should be highly appreciated in the many applications for which real-time segmentation is required, such as the autonomous deployment of sensor nodes in a given environment or the detection of flaws in quality inspection of materials. Moreover, FODPSO is slightly faster than DPSO because its fractional calculus is able to control the convergence rate of the algorithm. A swarm behavior can be divided into two main activities: exploitation and exploration. The former controls the convergence of the algorithm, thus allowing a good short-term performance. However, if the exploitation level is too high, then the algorithm may get stuck on local solutions. The latter, however, controls the diversification of the algorithm, which allows exploring new solutions, thus improving the long-term performance. However, if the exploration level is too high, then the algorithm may take too much time to find the global solution. In the DPSO, the trade-off between exploitation and exploration can only be controlled by adjusting the inertia weight. A large inertia weight improves exploration activity, however, the exploitation may be improved using a small inertia weight. Because the FODPSO presents a fractional-calculus strategy to control the convergence of particles with memory effect, the coefficient α allows providing a higher level of exploration while ensuring the global solution of the algorithm (Ghamisi et al. 2012).

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.