Particle swarm optimization with adaptive restriction factor

Optimization has been a technique that has been used in a wide variety of areas in recent years. Computer networks, economics, image processing, robotics and many other areas are used. The Particle Swarm Optimization (PSO) method which is developed from standard PSO, inspired by bird and fish swarms, is a fast converging algorithm. In this study, K and w ,generally used as static, are decreased with as aritmethic, geometric and inversely inspired by temperature parameter in Simulated Annealing and the performances of the algorithms are observed on benchmark function. The results that obtained with constant parameters, compared with reducing methods that applied and it has been shown that the inverse function gives more successful results than the fixed parameter for unconstrained optimization test problems. As can be seen in the result graphs, tests made with functions decreasing in PSOs using inertia weight (w) and adaptive constraint parameter (K) yielded better results than tests with constant values. In PSO algorithms developed for real-time applications, it is suggested that parameters can be used reducing functionally instead of constant use of parameters.