摘要
本文提出了一种群体启发进化规划(Population Heuristic Evolutionary Programming,PHEP)方法,在进化过程中,通过群体的四个参数,把握群体中个体的分布情况,并通过这些信息有效地调整个体的变异步长,使用标准测试问题对PHEP方法进行了测试和比较,试验数据表明:PHEP方法能够较好地平衡群体压力和选择压力,搜索到的结果要明显优于先前的一些方法。
This paper presents a new method for optimization problem which named population heuristic evolutionary programming (PHEP), it can grasp the information of distribution-status of population by control four parameters of population in the evolution process. PHEP adjusts the mutation size of individual according to such information. Ex- periments for benchmark problem imply that the PHEP can keep the balance of the population press and the selection press very well, the results are great better than other version EP methods.
出处
《计算机科学》
CSCD
北大核心
2005年第7期145-147,167,共4页
Computer Science
基金
国家自然科学基金(批准号:60175024)
教育部"符号计算与知识工程"重点实验室资助