摘要
针对鲸鱼优化算法难以跳出局部最优导致收敛精度不足的问题,提出一种融合了T-分布小波变异和多项式差分学习策略的鲸鱼优化算法.该算法首先引入Circle混沌扩大搜索范围,提高收敛速度;然后采用T-分布小波变异策略平衡全局和局部搜索能力;最后采用多项式差分学习策略改进算法的优化精度.对3种改进策略作单一引入的仿真对比分析,并将改进的鲸鱼优化算法在12个可变维度的基准测试函数上进行仿真,对本文改进的鲸鱼优化算法与其他改进策略的鲸鱼优化算法以及其他几种智能算法进行比较.结果表明,基于T-分布小波变异和多项式差分学习策略的改进鲸鱼优化算法具有较好的稳定性,收敛速度和精度更好.
A whale optimization algorithm based on T-distributed wavelet variation and polynomial difference learning strategy is proposed for the problem of the less accurate convergence of the whale optimization algorithm.The algorithm first introduces the circle chaos to expand the search scope and improve the convergence speed.Then,the T-distributed wavelet variation strategy balanced the search capability of the global and the bureau.Finally,the optimization precision of the polynomial difference learning strategy improved algorithm was adopted.In this paper,the simulation of three kinds of improvement strategies is compared and analyzed,and the improved whale optimization algorithm is simulated by the benchmark test function of 12 variable dimensions,and the whale optimization algorithm of this paper is compared with the whale optimization algorithm and several other intelligent algorithms of his improved strategy.The results show that the improved whale optimization algorithm based on T-distribution wavelet variation and polynomial difference learning strategy has good stability,convergence speed and precision.
作者
毛清华
赵冰
王迎港
MAO Qinghua;ZHAO Bing;WANG Yinggang(School of Economics and Management,Yanshan University,Qinhuangdao 066004,China)
出处
《小型微型计算机系统》
CSCD
北大核心
2024年第10期2362-2369,共8页
Journal of Chinese Computer Systems
基金
国家重点研发计划子项目(2020YFB1712802)资助.
