Embodied evolution versus cooperative coevolution in multi-robot optimization

Abstract

In this paper we show the potentiality of Embodied Evolution in the optimization of general multi-robot systems, as compared to state-of-the-art approaches based on Cooperative Coevolution. The comparison is carried out in a real application problem of coordinating a team of autonomous UAVs for surveillance which, in parallel, must optimize their location accuracy. The results show the advantages of using Embodied Evolution algorithms to notably reduce the number of evolution steps while maintaining the performance level.