This paper proposes the realization of evolutionary computation by chaotic dynamics instead of conventional GA (Genetic Algorithm) for such intelligent agents as welfare robots which assist humans. This evolutionary computation is realized by applying chaotic retrieval on associative memories. We apply this evolutionary computation to multiagent robots which move abraeast. Essentially the process of this evolutionary computation is parallel processing. Therefore, we implement its parallel processing algorithm on A-NET (Actors NETwork) parallel object-oriented computer, and show the usefulness of parallel processing for proposed evolutionary computation.
展开▼
