基于三支决策的多目标优化自然计算策略研究  

Research on natural computing strategy for multi-objective optimization based on three-way decision

作　　者：张心茹 季伟东[1] 岳玉麒 殷曾祥 ZHANG Xinru;JI Weidong;YUE Yuqi;YIN Zengxiang(College of Computer Science and Information Engineering,Harbin Normal University,Harbin 150025,China)

机构地区：[1]哈尔滨师范大学计算机科学与信息工程学院,哈尔滨150025

出　　处：《智能计算机与应用》2023年第2期134-138,共5页Intelligent Computer and Applications

基　　金：国家自然科学基金(31971015);黑龙江省自然科学基金(LH2021F037)。

摘　　要：为提高优化算法的优化效率,解决早熟收敛的问题,本文提出一种基于三支决策的多目标优化自然计算策略。利用分段Tent混沌初始化种群,生成均匀分布的初始化种群;引入三支决策思想,根据适应度值大小将种群分为正域、负域以及边界域,分别对三域中的最优个体执行不同的变异行为;结合个体在目标空间中的欧氏距离,充分发掘可能为最优解的潜在价值。将该策略分别应用到粒子群算法及灰狼算法中,并与这两个经典算法进行对比,实验结果表明:该策略具有更好的求解精度和更快的收敛速度,具有较高的寻优性能以及一定的普适性。In order to enhance the efficiency of the optimization algorithm and solve the problem of early convergence,a natural computational strategy of multi-objective optimization based on three-way decision is proposed.Using segmented Tent chaotic initialization population,a uniformly distributed initialization population is generated.The idea of three-way decision is introduced to divide the population into positive domain,negative domain and boundary domain according to the size of fitness value,and different variational behaviors are performed for the optimal individuals in the three domains respectively.The Euclidean distance of individuals in the target space is combined to fully explore the potential values that may be optimal solutions.The strategy is applied to the particle swarm optimization algorithm and the grey wolf optimization algorithm,respectively.The experimental results show that the strategy has better solution accuracy,faster convergence speed and higher performance of finding the optimal value as well as certain universality.

关 键 词：三支决策 多目标优化 自然计算 粒子群算法 灰狼算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]