Parallel heterogeneous genetic algorithms for continuous optimization

In this paper we address the physical parallelization of a very efficient genetic algorithm (GA) known as gradual distributed real-coded GA (GD-RCGA). This search model naturally provides a set of eight sub-populations residing in a cube topology having two faces for promoting exploration and exploitation. The resulting technique has been shown to yield very accurate results on continuous optimization by using crossover operators tuned to exploit and explore the space inside each sub-population. Here, we encompass the first actual parallelization of the technique, and get deeper into the importance of running a synchronous versus an asynchronous version of the basic GD-RCGA model. Our results indicate that this model maintains a very high level of accuracy for continuous optimization when run in parallel, as well as we show the similarities between the sync and async versions. Finally, we show that async parallelization is really more scalable than the sync one, suggesting future research lines for WAN execution and new models of search based in the two-faced cube of the original model.