An Improved Particle Swarm Optimization Algorithm Based on S-shaped Activation Function for Fast Convergence

In this paper an Improve particle swarm optimization algorithm(IPSO) is proposed in which an S-shaped activation function, which is inspired by the neural networks is used to update the acceleration factors, which play a significant role in fast convergence of particles within a given search space. So, in order to keep parity between the exploration and exploitation this S-shaped activation function take into account both the distance of the particle to its Pbest(Personal best position) and from a particle to its Gbest(Global best position), That's how its enhance the convergence rate. The proposed Improved PSO algorithm with S- shaped activation function is tested on some famous complex benchmark functions and its is also compared with some well-known PSO variants.