New mechanisms to enhance the performances of an adaptive algorithm of Particle Swarm Optimization

摘要

The aim of this paper is to present an improvement of the multiobjective TRIBES (MO-TRIBES). The main idea of this improvement is to propose two new operators: a mutation, which is applied to good particles and four processes of resets, which are applied to bad particles. The aim of the integration of those mechanisms is to insure a good exploration and/or exploitation of the search space. Besides, in our study, we proposed different percentages to apply these operators. The mechanisms proposed are validated using ten different functions from specialized literature of multi-objective optimization. The obtained results show that using these operators is valid as it is able to improve the quality of the solutions in the majority of case.