An Improved Particle Swarm Algorithm with Heuristic Factors and Empirical Operators

Abstract

An improved particle swarm algorithm (PSA) is proposed based on the universal framework of the intelligence optimization algorithms (IOA). Regarded as increment to the standard PSA, the improved PSA takes the essence of the standard PSA, and additionally utilizes various heuristic factors and different empirical operators introduced by multiple other IOA. The capability of global and local optimization can be strengthened by the free search, the differential evolution and the quantum rotation comprehensively. The genetic variation is adopted to avoid premature convergence, while the greed strategy is adopted to guarantee monotonous convergence. The numerical simulation indicates that the improved PSA achieves better performance than the standard PSA.