摘要
首先提出一个刻画进化种群多样性的函数 .考虑种群多样性与进化代数的关系 ,提出一个作用函数 .在此基础上 ,提出一种自适应遗传算法 ,该算法中交叉和变异操作的点数随种群多样性函数和作用函数而变化 .多峰值函数优化结果表明 ,该算法可以有效地解决早熟收敛问题 。
The function scaling diversity of evolution population was put forth. Taking the relation between diversity of evolution population and evolution time into account an effect function was present. Based on it, a novel adaptive genetic algorithm was put forth, in which the points of crossover and mutation operators change with diversity of evolution population and effect factor. The optimization result of multi-peak value function shows that the algorithm presented can solve premature convergence problem effectively and converge to the globally optimal solution.
出处
《中国矿业大学学报》
EI
CAS
CSCD
北大核心
2003年第1期68-70,共3页
Journal of China University of Mining & Technology
关键词
遗传算法
早熟收敛
作用函数
交叉
变异
genetic algorithm
premature convergence
effect factor
diversity of evolution population
