Modifications of Differential Evolution with Composite Trial Vector Generation Strategies
Modifications of Differential Evolution with Composite Trial Vector Generation Strategies
复制标题
DOI：
10.1007/978-3-642-32922-7_12
复制
发表时间：
2012
期刊：
影响因子：
--
通讯作者：
J. Tvrdík
中科院分区：
文献类型：
--
作者：
J. Tvrdík
最近提出了具有复合试验向量生成策略和控制参数的差分进化算法。与最先进的差分进化变体相比，该算法的性能据称更好或更具竞争力。当我们尝试根据已发布的描述来实现该算法时，出现了几个修改后的变体，这些变体遵循该算法的描述。在基准问题中对算法的这些变体进行了实验比较。新提出的变体之一显着优于其他变体，包括算法作者在其发表的实验比较中使用的变体。
Differential evolution algorithm with composite trial vector generation strategies and control parameters has been proposed recently. The performance of this algorithm is claimed to be better or competitive in comparison with the state-of-the-art variants of differential evolution. When we attempted to implement the algorithm according to the published description, several modified variants appear to follow the description of the algorithm. These variants of the algorithm were compared experimentally in benchmark problems. One of newly proposed variants outperforms the other variants significantly, including the variant used by the authors of the algorithm in their published experimental comparison.