灰狼优化与差分进化的混合算法及函数优化

发布时间：2018-05-21 15:54

本文选题：优化算法 + 混合优化算法　；参考：《计算机科学》2017年09期

【摘要】：灰狼优化(Grey Wolf Optimization,GWO)算法是近年被提出的一种新型智能优化算法,具有收敛速度快和优化精度高的特点,但对于一些复杂优化问题易陷入局部最优。差分进化(Differential Evolution,DE)算法的全局搜索能力强,但其性能对参数敏感,且局部搜索能力不足。为了发挥二者各自的优点并弥补存在的缺陷,提出了一种灰狼优化与差分进化的混合优化算法。首先使用嵌入趋优算子的GWO算法搜索,以便在更短的过程中获得更高的优化精度和更快的收敛速度;然后采用自适应调节参数的差分进化策略来进一步提高算法对复杂优化函数的寻优性能,从而获得一种高性能的混合优化算法,以便能更高效地解决各种函数优化问题。对12个高维函数的优化结果表明,与标准GWO,ACS,DMPSO及SinDE相比,新的混合优化算法不仅具有更好的收敛速度和优化性能,而且具有更好的普适性,更适用于解决各种函数优化问题。
[Abstract]:Grey Wolf Optimization (GWO) algorithm is a new intelligent optimization algorithm proposed in recent years. It has the characteristics of fast convergence and high precision, but it is easy to fall into local optimization for some complex optimization problems. Differential evolution (DEE) algorithm has strong global search ability, but its performance is sensitive to parameters and its local search ability is insufficient. In order to give full play to their respective advantages and make up for the defects, a hybrid optimization algorithm of gray wolf optimization and differential evolution is proposed. First, the GWO algorithm with embedded optimization operator is used to search in order to obtain higher optimization accuracy and faster convergence speed in a shorter process. Then the differential evolution strategy of adaptive parameter adjustment is adopted to further improve the optimization performance of the algorithm for complex optimization functions, so as to obtain a high performance hybrid optimization algorithm, so as to solve all kinds of function optimization problems more efficiently. The optimization results of 12 high-dimensional functions show that the new hybrid optimization algorithm not only has better convergence speed and optimization performance, but also has better universality, and is more suitable for solving various function optimization problems compared with the standard GWO ACSS-DMPSO and SinDE.
【作者单位】：河南师范大学计算机与信息工程学院;河南省高校计算智能与数据挖掘工程技术研究中心;
【基金】：河南省重点科技攻关项目(132102110209) 河南省基础与前沿技术研究计划项目(142300410295)资助
【分类号】：TP18

【相似文献】