计算机工程与应用 ›› 2021, Vol. 57 ›› Issue (18): 86-96.DOI: 10.3778/j.issn.1002-8331.2008-0178

• 理论与研发 • 上一篇    下一篇

动态反向搜索更新位置的改进灰狼优化算法

王梦璐,李连忠   

  1. 江南大学 理学院,江苏 无锡 214122
  • 出版日期:2021-09-15 发布日期:2021-09-13

Improved Grey Wolf Optimization Algorithm Based on Dynamic Reverse Search for Updated Position

WANG Menglu, LI Lianzhong   

  1. College of Science, Jiangnan University, Wuxi, Jiangsu 214122, China
  • Online:2021-09-15 Published:2021-09-13

摘要:

针对灰狼优化算法(GWO)后期收敛速度慢、易陷入局部最优等问题,提出一种动态反向搜索更新位置的改进灰狼优化算法(DAGWO)。该算法在原始的位置更新公式中引入个体历史最优位置引导策略,以加快算法的收敛速度;同时,引入反向搜索因子,该因子依据种群早熟判别指标动态调节自身取值,在算法陷入局部极值时令灰狼个体向整个种群中最差个体方向进行反向搜索,以提高种群跳出局部极值的能力。此外,构造了一种新型局部扰动的非线性收敛因子[a],以平衡算法的全局和局部搜索能力。对20个经典测试函数进行仿真实验,结果表明在求解精度、收敛速度和算法的稳定性上,DAGWO算法与标准智能优化算法和其他相关改进算法相比更有优越性。

关键词: 改进灰狼优化算法, 个体历史最优位置, 早熟判别指标, 反向搜索因子, beta随机调整数

Abstract:

Aiming at the disadvantages of Grey Wolf Optimization(GWO), such as easy to fall into local optimization and slow convergence speed, an improved GWO algorithm based on dynamic reverse search for updated position(DAGWO) is proposed. In order to accelerate the convergence speed of the algorithm, the optimal position guidance strategy of individual history is introduced into the original position updating formula. At the same time, a reverse search factor is introduced, which dynamically adjusts its value according to the early-maturity discrimination index of the population. When the algorithm falls into the local optimum, gray wolf individuals are reversely searched to the direction of the worst individual in the whole population, so as to improve the ability of the population to jump out of the local optimum. In addition, a novel nonlinear convergence factor for local perturbation is constructed to balance the global and local search capabilities of the algorithm. The simulation results of 20 classical test functions show that the DAGWO algorithm is superior to the standard intelligent optimization algorithm and other related improved algorithms in solving precision, convergence speed and stability of the algorithm.

Key words: improved grey wolf optimization algorithm, the best position in individual history, early maturity index, reverse search factor, beta random adjustment