PSO Poem Analysis - 725 Words

There are several complexes as well as nonlinear problems which can be solved using a special type of optimization techniques. Some of which are inspired by natural phenomenon like Fish Schooling, Swarm behavior, Ant colony etc. Among these, the PSO is a simple but powerful search technique which gives better results. The PSO algorithm maintains a population of particle where each particle represents a potential solution to an optimization problem, each potential solution in PSO is also associated with a randomized velocity and then the potential solution called particles are flown through the problem space. Let S be the size of swarm where each particle i can be represented as an object with several characteristics. A population of the†¦show more content†¦The update position is updated only if current position X_id^((t+1)) has better results compared to the previous position X_id^((t)) otherwise, the position of the particle remains unchanged as the new position of a parti cle depends on previous position, pbest and gbest. Therefore pbest and gbest play an important role in optimization technique [4,8]. 2.1.1 PSO Variants The two main feature of optimization algorithms is exploitation and exploration. Exploration is an ability of the algorithm to explore a different region of search space in a manner to locate good optimum while exploitation is the ability to concentrate the search around the promising area in order to refine a candidate solution. Also, exploration deals with global optimal and exploitation provide local optimization and to handle these feature inertia weight parameter (w) is used in modified PSO. A Higher value of ‘w’ gives better exploration and lower value of ‘w’ gives better exploitation [4,9]. 2.1.2 Constant Weighted Inertia PSO (CWI - PSO) CWI –PSO provides a constant rate for Exploitation and Exploration. Here inertia weight remains constant in all iteration. This algorithm is problem specific and it is difficult to find out the constant value for a particular problem. To find constant inertia weight many simulations are carried out with different inertia weights. It is defined as wt=c