El-Desouky N., Ghali N., Zaki M.
Faculty of Science (for girls)- Al-Azhar University- Cairo – Egypt
In this paper we propose a new approach to weight variation in Particle Swarm Optimization
(PSO) called Exponential Particle Swarm Optimization (EPSO). EPSO depends on exponential
variation for the inertia weight; which under all testing cases makes the particle swarm converges
very quickly towards the optimal positions compared with the linear particle swarm optimization
invented by Kennedy and Eberhart 1995. This comparison is based on different types of
benchmark functions, with asymmetric initial range settings, which are selected as test functions.
The proposed EPSO increases the possibility to find the optimal positions as it decrease the
number of failure. Also the experimental results show that the proposed technique decreases the
number of iterations needed to reach the optimal solution by ratio varying from 77% to 96%
depending on the tested optimization problem.