摘要
在分析了单亲遗传算法的优越性与存在不足的基础上,借鉴生物免疫概念与理论并结合并行计算的思想,提出了一种新的遗传算法——基于免疫的并行单亲遗传算法。理论分析和仿真结果表明,该算法不仅能够有效地保持群体多样性,而且减轻了遗传算法后期的波动现象,同时收敛速度有明显的提高。
A novel algorithm parallel partheno genetic algorithm based on the immune programming （IPPPGA） was proposed, the algorithm analogies to the concept and theory of immunity in biotic science and unifies the parallel computation thought, after analyzing the advantages and disadvantages of the partheno genetic algorithm. The theory analysis and experimental results show that the algorithm not only effectively keeps population diversity, but also alleviates the undulate phenomenon in evolutionary latter stages, meanwhile increases the convergent speed and computational efficiency greatly.
出处
《通信学报》
EI
CSCD
北大核心
2007年第8期99-104,共6页
Journal on Communications
基金
山西省青年科技研究基金资助项目(20031008)~~
关键词
并行
单亲遗传算法
免疫算法
非均匀变异算子
免疫算子
parallel
partheno genetic algorithm
immune programming
nonuniform mutation operator
immune operator
