Combination of Novelty Search and Fitness-Based Search Applied to Robot Body-Brain Co-Evolution

Evolutionary algorithms are a frequently used technique for designing morphology and controller of a robot. However, a significant challenge for evolutionary algorithms is premature convergence to local optima. Recently proposed Novelty Search algorithm introduces a radical idea that premature convergence can be avoided by ignoring the original objective and searching for any novel behaviors instead. In this paper, we apply novelty search to the problem of body-brain co-evolution. We show that novelty search significantly outperforms fitness-based search in a deceiving barrier avoidance task. Furthermore, we demonstrate an unexpected result that switching from novelty search to fitness-based search after the deceptive barrier is overcome does not significantly improve overall search performance.