Aiming at the two characteristics of premature convergence of particle swarm optimization that the particle velocity approaches 0 and particle swarm congregate, this paper learns from the annealing function of the simulated annealing algorithm and adaptively and dynamically adjusts inertia weights according to the velocity information of particles to avoid approaching 0 untimely. This paper uses the good uniformity of Anderson chaotic mapping and performs chaos perturbation to part of particles based on the information of variance of the population’s fitness to avoid the untimely aggregation of particle swarm. The numerical simulations of five test functions are performed and the results are compared with several swarm intelligence heuristic algorithms. The results shows that the modified algorithm can keep the population diversity well in the middle stage of the iterative process and it can improve the mean best of the algorithm and the success rate of search.
The Particle Swarm Optimization Algorithm with Adaptive Chaos Perturbation
2016-10-17
doi:10.15837/ijccc.2016.6.2525
INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL; Vol 11 No 6 (2016): International Journal of Computers Communications & Control (December); 804-818 ; 1841-9844 ; 1841-9836 ; 10.15837/ijccc.2016.6
Article (Journal)
Electronic Resource
English
DDC: | 629 |
Improved particle filter algorithm based on chaos particle swarm optimization
British Library Online Contents | 2016
|An Adaptive Particle Swarm Optimization Algorithm with Local Search
British Library Online Contents | 2015
|Particle Swarm Optimization Algorithm
British Library Online Contents | 2005
|A Particle Swarm Optimization Algorithm with Initialized Adaptive Inertia Weights
British Library Online Contents | 2012
|Adaptive Range Particle Swarm Optimization
British Library Conference Proceedings | 2006
|