Quantitative Evaluation of Optimization Efficiency for Genetic Algorithms
-
摘要: 为了评价遗传算法的优化效率,提出了"平均截止代数"和"截止代数分布熵"的概 念,并用二者组成的平面测度作为评价准则.在此基础上,以浮点型遗传算法为例,对不同遗 传算子的优化效率进行了详细的研究.结果表明,不同遗传算子对应着不同的优化效率;这为 选择高效的遗传算子提供了科学依据.Abstract: In order to evaluate the optimization efficiency of genetic algorithms, this paper presents two indices--"the average truncated generation" and "the distribution entropy of truncated generations". Thereafter, they are unified as a monolithic criterion. Based on this criterion, the optimization efficiencies of several genetic operators are investigated in detail with the example of float-type genetic algorithms. The results show that the variation of genetic operators corresponds to that of optimization efficiencies. The conclusion provides a scientific basis for selecting the efficient genetic operators.
-
Key words:
- Genetic algorithms /
- optimization /
- efficiency
计量
- 文章访问数: 2274
- HTML全文浏览量: 93
- PDF下载量: 976
- 被引次数: 0