摘要
基于并行进化种族间的协作和竞争机制，给出了ＣＰＣＡ算法。新算法的进化操作更符合自然选择机制，在动态增加新种族的同时，亦动态删除老化、竞争力弱的种族。该算法用来进化ＢＰ网，经实验证明能有效地提高解的质量，并降低进化时间。
This paper proposes a new parallel genetic algorithm based on competition and cooperation of species. The new algorithm is more natural in that it can dynamically create new species and delete weak species.A BP network is taken as an example to examine the algorithm. The simulated results demonstrate that the algorithm′s solution quality and coverage rate are much better than those of the conventional techniques.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
1999年第3期21-25,共5页
Journal of Hefei University of Technology：Natural Science
基金
国家自然科学基金
