Multiple Particle Swarm Optimizers with Diversive Curiosity

In this paper we propose a new method, called multiple particle swarm optimizers with diversive curiosity (MPSOα/DC), for improving the search performance of the convenient multiple particle swarm optimizers. It has three outstanding features: (1) Implementing plural particle swarms simultaneously to search; (2) Exploring the most suitable solution in a small limited space by a localized random search for correcting the solution found by each particle swarm; (3) Introducing diversive curiosity into the whole particle swarms to comprehensively deal with premature convergence and stagnation. To demonstrate the effectiveness of the proposed method, computer experiments on a suite of benchmark problems are carried out. We investigate the characteristics of the proposed method, and compare the search performance with other methods such as EPSO, OPSO, and RGA/E. The experimental results indicate that the search performance of MPSOα/DC is superior to EPSO, OPSO, and RGA/E for the given benchmark