Abstract
Swarm robotics is the research field of multi-robot systems which consist of many homogeneous autonomous robots without any type of global controllers. In this paper, an evolutionary robotics (ER) approach, i.e., the method that robot controllers represented using artificial neural networks are designed by evolutionary algorithms, is applied. For complex tasks in ER approach, however, all individuals in the first generations are often scored with the same null value, and as a consequence the selection process cannot operate. In order to overcome this "bootstrap" problem, we apply an incremental approach to evolution within the context of evolutionary robotics. As a benchmark of robotic swarms, cooperative food-foraging problems are conducted to examine their performance.
