Agent-based parallel Particle swarm optimization based on group collaboration

Particle swarm optimization (PSO) is a population-based evolutionary computation technique. The standard parallel Particle swarm optimization (PSO) is a promising scheme for solving NP-complete problems due to its fast convergence, fewer parameter settings and ability to fit dynamic environmental characteristics. A modified version of standard parallel particle swarm optimization algorithm named Agent-based Parallel PSO (PPSO) is presented to solve the large-scale optimization problem at a faster convergence rate. The new algorithm provides the information exchange among the particle sets which aims to accelerate the rate of convergence. Parallelization is based on the client-server model in which the particle set with global best value acts as a server and others are clients with agents in a single iteration. The agents contain the worst value information of respective particle set which they have to exchange with the global best value among particle sets in the swarm. The server is the center of data exchange, which deals with agents and manages the sharing of global best position among the individual clients by the replacement of worst position with global best position. The information exchange in particle sets has been done so that the new result will be the better as compared to previous result.

I. Introduction

The method of Particle Swarm optimization was introduced in 1995 by Kennedy and Eberhart [1] and has been successfully applied to several different problems, including the training of neural networks, structural and topology optimization, and image recognition. The underlying principle of this stochastic population-based method is that, a number of agents coordinate their search patterns in the design space by communicating the locations of promising regions.