LEARNING TO COMPOSE FUZZY BEHAVIORS FOR AUTONOMOUS AGENTS

We present S-ELF, an evolutionary algorithm that we have developed to learn the context of activation of fully logic controllers implementin g fuzzy behaviors for an autonomous agent. S-ELF learns context metaru les that are used to coordinate basic behaviors in order to perform co mplex tasks in a partially and imprecisely known environment Context m etarules are expressed in terms of positive and negated fuzzy predicat es. We also show how S-ELF can learn robust and portable behaviors, th us reducing the time and effort to design behavior-based agents. (C) 1 997 Elsevier Science Inc.