作者: Andrea Bonarini , Filippo Basso
DOI: 10.1016/S0888-613X(97)00002-9
关键词:
摘要: Abstract We present S-ELF, an evolutionary algorithm that we have developed to learn the context of activation fuzzy logic controllers implementing behaviors for autonomous agent. S-ELF learns metarules are used coordinate basic in order perform complex tasks a partially and imprecisely known environment. Context expressed terms positive negated predicates. also show how can robust portable behaviors, thus reducing time effort design behavior-based agents.