融合知识共享和精英反向学习的成长优化算法

Growth optimization algorithm integrating knowledge sharing and elite opposition-based learning

为解决成长优化(growth optimizer,GO)算法易陷入局部最优的问题,提出一种融合知识共享和精英反向学习的成长优化算法(growth optimization algorithm integrating knowledge sharing and elite opposition-based learning,KSOBLGO).该算法结合基于知识共享的优化算法(gaining-sharing knowledge based algorithm,GSK)和精英反向学习策略,旨在提高算法的全局搜索能力并避免陷入局部最优解的困境.首先采用精英反向学习策略,通过引入优秀个体的信息增加初始种群的多样性,避免算法过早收敛,导致局部滞留.将基于知识共享的优化算法融合其中,称为知识共享阶段,该阶段提高种群中较优个体的局部搜索能力和较差个体的全局搜索能力,使探索和开发达到良好的平衡状态.经过对13个基准测试函数进行仿真实验,研究结果表明KSOBLGO算法在收敛速度和寻优精度等方面取得显著的提升,验证该改进算法的有效性.

To address the issue of the Growth Optimizer(GO)algorithm getting trapped in local optima,this paper proposes a new variant of the growth optimization algorithm called KSOBLGO.The KSOBLGO algorithm combines the Gaining-Sharing Knowledge Based Algorithm(GSK),which is a knowledge-sharing optimization algorithm,with the elite backward learning strategy to improve the algorithm's global search capability and avoid getting stuck in local optima.The algorithm first employs the elite backward learning strategy to increase the diversity of the initial population by introducing information from outstanding individuals.This helps to prevent the algorithm from converging prematurely,thus avoiding local stagnation.The algorithm incorporates the knowledge-sharing optimization algorithm,referred to as the knowledge-sharing phase,which enhances the local search ability of superior individuals in the population and the global search ability of inferior individuals.This achieves a balance between exploration and exploitation.Simulations and experiments were conducted on 13 benchmark test functions.The research results demonstrate significant improvements in convergence speed and optimization accuracy achieved by the KSOBLGO algorithm,validating its effectiveness as an enhanced algorithm.