An improved particle swarm optimization algorithm for global optimizations of electromagnetic devices

摘要

Based on comprehensive simulations of bird flocking or fish schooling, an Improved Particle Swarm Optimization algorithm (IPSO) is proposed in this paper. The improvements include mainly the introduction of a new generating mechanism for initial particles to improve the convergence performances, the restarting strategy to avoid a stagnation phenomenon, and the design of novel velocity and position updating formulae to enhance the global search ability of the available PSOs. The numerical results on both a mathematical function and the TEAM Workshop problem obtained by using different optimization algorithms are reported and the performances are compared. Our numerical results suggest that the proposed IPSO algorithm is superior to its precursors in sense of the global search ability.