Empirical Study of Simultaneous Perturbation Particle Swarm Optimization

In this paper, we propose some different optimization schemes which are combinations of the particle swarm optimization and the simultaneous perturbation optimization method. The proposed schemes can utilize local information of an objective function and global shape of the function at the same time. These characteristics are from the simultaneous perturbation optimization method and the particle swarm optimization. The schemes have good properties of global search and efficient local search capability. Moreover, the schemes themselves are very simple and easy to implement. These methods only require values of the function similar to the original particle swarm optimization and the simultaneous perturbation method. The proposed schemes are investigated using some test function to know convergence properties such as convergence rate or convergence speed.