New Particle Swarm Optimization with Diminishing Population

Particle Swarm Optimization (PSO) is a method of social investigation which its function is on this principle that in every moment, any particle regulates its position in searching space regarding to best resting position and best position in its neighbouring. Regarding to chronological process when the number of local minimum points as fitness function would be high, PSO algorithm in which will be easily captured by value of local optimum. Hence in this paper it is presented a method for implementation of PSO algorithm in which regarding to worst place of each particle and diminishing population by removing of low operation particles, by inhibition of capturing local optimum amounts and drives the particles toward the successful regions. The results show that implementation of this method for function with high local minimum would cause general searching, decreases the number of calculations and would result better optimum value than to PSO.