Comparison study of the combination of the SPSA algorithm and the PSO algorithm
Authors:
Bertrand Ngansop,
Stefan Götz,
Martin Eckl
Abstract:
Particle swarm optimization (PSO) is attracting an ever-growing attention and more than ever it has found many application areas for many challenging optimization problems. It is, however, a known fact that PSO has a severe drawback in the update of its global best (gbest) particle, which has a crucial role of guiding the rest of the swarm. In this paper, we propose three efficient solutions to re…
▽ More
Particle swarm optimization (PSO) is attracting an ever-growing attention and more than ever it has found many application areas for many challenging optimization problems. It is, however, a known fact that PSO has a severe drawback in the update of its global best (gbest) particle, which has a crucial role of guiding the rest of the swarm. In this paper, we propose three efficient solutions to remedy this problem using the SPSA Algorithm. In the first approach, gbest is updated with respect to a global estimation of the gradient and can avoid getting trapped into a local optimum. The second approach is based on the formation of an alternative or artificial global best particle, the so-called aGB, which can replace the native gbest particle for a better guidance, the decision of which is held by a fair competition between the two. The third approach is based on the update of the swarm particle. For this purpose we use simultaneous perturbation stochastic approximation (SPSA) for its low cost. Since SPSA is applied only to the gbest (not to the entire swarm) or to the entire swarm, both approaches result thus in a negligible overhead cost for the entire PSO process. Both approaches are shown to significantly improve the performance of PSO over a wide range of non-linear functions, especially if SPSA and PSO parameters are well selected to fit the problem at hand. As in the basic PSO application, experimental results show that the proposed approaches significantly improved the quality of the Optimization process as measured by a statistic analysis.
△ Less
Submitted 25 April, 2022;
originally announced April 2022.
