›› 2016, Vol. 29 ›› Issue (1): 25-.

• 论文 • 上一篇    下一篇

梯度策略自适应差分进化算法

杨俊,魏静萱   

  1. 差分进化算法是一种有效求解全局优化问题的方法,为进一步提高求解精度,加快求解过程,文中提出一种梯度策略自适应差分进化算法。该算法是在差分进化算法中加入梯度下降法,使其不仅有较好的全局搜索能力,且具有传统优化方法的快速局部搜索能力,因此具有较高搜索精度和较快的搜索过程。通过对CEC2005测试集中的1~14号测试函数进行仿真实验,并与SaDE,NSDE以及CMAES等算法实验结果进行了对比,结果表明了该算法的有效性。
  • 出版日期:2016-01-15 发布日期:2016-02-25
  • 作者简介:魏静萱(1981—),女,博士,副教授,硕士生导师。研究方向:智能计算。杨俊(1992—),男,硕士研究生。研究方向:智能计算。
  • 基金资助:

    国家自然科学基金资助项目(61203372)

Adaptive Differential Evolution Algorithm Based on Gradient Search Strategy

YANG Jun,WEI Jingxuan   

  1. The differential evolutionary algorithm is effective in solving global optimization.An improved adaptive differential evolution algorithm based on gradient search strategy is proposed to improve the accuracy of the solved solutions.The improved gradient search strategy is introduced into differential evolutionary algorithm to solve large scale optimization problems.The proposed algorithm is capable of both global and local search.The simulation results show that the proposed algorithm has better results compared to SaDE,NSDE and CMAES for benchmark functions 1~14 in CEC 2005.
  • Online:2016-01-15 Published:2016-02-25

摘要:

差分进化算法是一种有效求解全局优化问题的方法,为进一步提高求解精度,加快求解过程,文中提出一种梯度策略自适应差分进化算法。该算法是在差分进化算法中加入梯度下降法,使其不仅有较好的全局搜索能力,且具有传统优化方法的快速局部搜索能力,因此具有较高搜索精度和较快的搜索过程。通过对CEC2005测试集中的1~14号测试函数进行仿真实验,并与SaDE,NSDE以及CMAES等算法实验结果进行了对比,结果表明了该算法的有效性。

关键词: 差分进化算法, 全局优化, 梯度下降法

Abstract:

The differential evolutionary algorithm is effective in solving global optimization.An improved adaptive differential evolution algorithm based on gradient search strategy is proposed to improve the accuracy of the solved solutions.The improved gradient search strategy is introduced into differential evolutionary algorithm to solve large scale optimization problems.The proposed algorithm is capable of both global and local search.The simulation results show that the proposed algorithm has better results compared to SaDE,NSDE and CMAES for benchmark functions 1~14 in CEC 2005.

Key words: differential evolution algorithm;global optimization;gradient descent

中图分类号: 

  • TP306.1