Cosine function applied to the inertia control in the Particle Swarm Optimization

This work presents a new mechanism to reduce statistically the chances of the optimization process of nonlinear functions stagnating in local minima, using the meta-heuristic Particle Swarm Optimization. Such mechanism adopts a nonmonotonic way to control the particle inertia, which is one of the factors responsible for this movement during the optimization process. For this, the cosine function was used as a basis for generating this behavior non-monotonic of inertia. Two ways to use the cosine function have been proposed, one maintaining its default behavior, and another using a kind of mirroring in the original cosine function. The experimental results of the methods used to inertia control were compared to the PSO original model aiming to show the potential to find a better solution related to the benchmark functions for complex problems.