摘要
针对蝴蝶优化算法存在种群多样性差、寻优精度低、收敛速度慢的不足,提出了拉丁超立方抽样的自适应高斯小孔成像蝴蝶优化算法。首先利用拉丁超立方抽样种群初始化策略以提高种群的多样性,从而增强算法的全局搜索能力;然后引入在不同进化时期自动调节搜索范围的自适应最优引导策略,平衡算法的全局和局部搜索能力,从而提升算法的寻优精度;最后采用高斯小孔成像策略,对最优个体进行扰动,使得种群个体向最优个体靠近,以进一步提升算法的寻优精度并加快算法的收敛速度。通过对14个基准测试函数进行仿真实验以及Wilcoxon秩和检验,结果表明改进算法的寻优精度、收敛速度、稳定性和可扩展性等性能均得到了较大提高。
Aiming at the shortcomings of Butterfly optimization algorithm,such as poor population diversity,low optimization accuracy and slow convergence speed,this paper proposed a self-adaptive Gaussian keyhole imaging butterfly optimization algorithm based on Latin hypercube sampling.Firstly,it used a Latin hypercube sampling population initialization strategy to enhance the population diversity and thereby improve the overall search ability of the algorithm.Then,it introduced the self-adaptive optimal guidance strategy,which could dynamically adjust the search range in different evolutionary periods,to balance the global and local search capabilities and hence improve the optimization accuracy of the algorithm.Finally,it adopted a Gaussian keyhole imaging strategy to disturb the optimal individuals,making the individuals of the population moving close to the optimal individuals,so as to further improve the solution accuracy and speed up the convergence of the algorithm.Through simulation experiments and Wilcoxon rank sum tests using 14 benchmark functions,the results show that the performance of the improved algorithm is greatly enhanced in terms of optimization accuracy,convergence speed,stability and scalability.
作者
徐杰
鲁海燕
赵金金
侯新宇
卢梦蝶
Xu Jie;Lu Haiyan;Zhao Jinjin;Hou Xinyu;Lu Mengdie(School of Science,Jiangnan University,Wuxi Jiangsu 214122,China;Wuxi Engineering Technology Research Center for Biological Computing,Jiangnan University,Wuxi Jiangsu 214122,China)
出处
《计算机应用研究》
CSCD
北大核心
2022年第9期2701-2708,2751,共9页
Application Research of Computers
基金
国家自然科学基金资助项目(61772013,61402201)
江苏省青年基金资助项目(BK20190578)。
关键词
蝴蝶优化算法
拉丁超立方抽样
自适应惯性权重
高斯小孔成像
高维优化
butterfly optimization algorithm
Latin hypercube sampling
self-adaptive inertia weight
Gaussian keyhole imaging
high dimensional optimization
