Authors: Luis M. Torres-Treviño
Addresses: Facultad de Ingeniería Mecánica y Eléctrica, Centro de Investigación e Innovación de Desarrollo de Ingeniería y Tecnología, Universidad Autónoma de Nuevo León, San Nicolás de los Garza, Nuevo León, México
Abstract: Swarm intelligence is a recent computational paradigm inspired in behaviour of social animals. The principal application of this paradigm is for optimisation problems where a near optimal solution is expected. Spatial position of members of the swarm is related with a set of potential solutions. Using an update rule related with specific species that have a swarm behaviour, fitness of every member of the swarm is increased. This update rule depends on a specific swarm; however, when this update rule is applied several times, the performance of the swarm is improved reducing the number of evaluations required to achieve a near optimal solution. This paper presents numerical studies on two swarm algorithms inspired in flock of birds and pack of wolfs where update rules are applied several times for global optimisation of six multivariable functions. The performance of the algorithm is improved significantly, inclusive in high dimension multimodal functions.
Keywords: swarm computation; swarm intelligence; implicit elitism; metaheuristics; global optimisation; exploration; exploitation; PSO; particle swarm optimisation; GWO; grey wolf optimisation.
International Journal of Bio-Inspired Computation, 2017 Vol.9 No.2, pp.65 - 76
Available online: 21 Mar 2017 *Full-text access for editors Access for subscribers Purchase this article Comment on this article