摘要
受人类进化过程的启发,提出了一种双系统协同进化的基因表达式编程算法DSCE-GEP。DSCE-GEP由自然进化系统和人工干预系统组成。人工干预系统包括个体干预和种群干预。个体干预是依据基因库对种群中的个体进行去劣和增优操作,旨在改善种群中个体的质量;种群干预通过引入随机和镜像个体来提高种群的多样性和全局寻优能力。与权威文献中改进的GEP关于函数发现问题的大量对比实验表明,本文算法在收敛速度、求解质量方面优于对比算法,具有明显的竞争力。
Inspired by the evolution process of human,we propose a double system co-evolutionary gene expression programming （DSCE-GEP）. The DSCE-GEP consists of a natural evolution system and an artificial intervention system. The latter includes individual intervention operation and population in- tervention operation. The individual intervention operation is based on a gene pool which aims to improve the quality of individuals. It has a repairing operator that removes the morbid genes in individuals and a strengthening operator that spreads eminent genes to the individuals. The population intervention opera- tion improves the diversity of population and the global searching ability of the algorithm by introducing a certain number of feasible random individuals and feasible mirror individuals. Experimental results show that the performance of the DSCE-GEP is better than that of other GEP algorithms proposed in re- lated literatures regarding function finding problems in terms of convergence speed and solution quality, and that it promises competitive performance.
出处
《计算机工程与科学》
CSCD
北大核心
2016年第11期2314-2320,共7页
Computer Engineering & Science
基金
国家自然科学基金(31170393)
陕西省自然科学基金(2012JM8023)
关键词
双系统协同进化基因表达式编程
基因库
人工干预系统
函数发现问题
double system co-evolutionary gene expression programming
gene pool
artificial intervention system
function finding problems
