基于近郊区和远郊区的果蝇优化新算法

发布时间：2018-05-02 10:04

  本文选题：果蝇优化算法 + 局部最优　； 参考：《计算机工程》2017年02期

【摘要】：在传统果蝇优化算法中,果蝇的新位置常被限定在特定区域内,因此,寻优结果对搜索半径依赖性强,导致算法极易陷入局部最优。为此,提出一种改进的果蝇优化算法。将果蝇在每个维度上的搜索范围分为2个部分,给出近郊区和远郊区的概念,引入局部最优导向因子,通过动态调整该因子协调果蝇在不同区域的搜索强度,通过随机选择果蝇位置向量中特定维度实现果蝇位置更新。仿真实验结果表明,与传统自适应混沌果蝇优化算法相比,该算法能有效避免搜寻半径的影响,且在收敛精度、收敛速度等方面具有明显优势。
[Abstract]:In the traditional Drosophila optimization algorithm, the new position of Drosophila is often confined to a specific region. Therefore, the search results are strongly dependent on the search radius, resulting in the algorithm is prone to fall into local optimization. Therefore, an improved algorithm for fruit fly optimization is proposed. The search range of Drosophila on each dimension is divided into two parts. The concepts of peri-suburb and far-suburb are given, and the local optimal guidance factor is introduced. The search intensity of Drosophila in different regions is coordinated by dynamically adjusting the factor. Drosophila position was updated by random selection of specific dimensions in the Drosophila position vector. The simulation results show that compared with the traditional adaptive chaotic Drosophila optimization algorithm, the algorithm can effectively avoid the influence of searching radius, and has obvious advantages in convergence accuracy and convergence speed.
【作者单位】： 中央财经大学信息学院;吉林大学计算机科学与技术学院;
【基金】：信息保障技术重点实验室开放基金(KJ-14-008)
【分类号】：TP18
，

本文编号：1833435