Overview of differential evolution algorithm and its improved algorithms
Overview of differential evolution algorithm and its improved algorithms
复制标题
DOI：
--
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
Xie Bai-qiao
中科院分区：
文献类型：
--
作者：
Xie Bai-qiao
差异进化(DE)算法是一种基于种群差异的进化算法，每个种群向量代表一个优化问题的候选解。该算法利用向量之间的差异来扰动整个种群来寻找最优解。综述了差异进化算法的基本原理、基准测试函数、优缺点，讨论了三角变异、基于对立的学习、均匀设计法等改进方法，并对未来的研究方向进行了展望。
Differential evolution(DE) algorithm is an evolution algorithm based on population differences.Each vector of population represents one candidate solution of the problem optimized.The algorithm finds the best result by perturbing the whole population with the differences between vectors.The basic principle,benchmark test functions,strong points and shortcomings of DE are reviewed.Some improved methods,such as trigonometric mutation,opposition-based learning,uniform design method,are discussed.And future research issues are also given.