Abstract

Differential Evolution (DE) is a powerful optimization procedure that self-adapts to the search space. Although it is well known that the population structure has a major influence on the behavior of DE, there are a few works studying its effect in DE algorithms. In this paper, an ensemble of population topologies with the DE algorithm (EPTDE) is proposed. In EPTDE, a pool of topology strategies exists throughout the evolution process to produce offspring. automatically updates the population topology to appropriate topology to avoid premature convergence. This method utilizes the information of the population effectively and enhance the performance of DE. The performance of EPTDE is evaluated on a set of benchmark problems and is compared with several DEs with different topologies. Experimental results demonstrate that EPTDE is better than, or at least comparable to, other DEs.

Paper Details