基于指数惯性权重和自适应变异的樽海鞘算法  

Salp swarm algorithm based on exponential inertia weight and adaptive mutation

作　　者：蔡艺君 贺兴时[1] 杨新社[2] CAI Yijun;HE Xingshi;YANG Xinshe(School of Science, Xi’an Polytechnic University, Xi’an 710048, China;School of Science and Technology, Middlesex University, London NW4 4BT, UK)

机构地区：[1]西安工程大学理学院,陕西西安710048 [2]密德萨斯大学科学与技术学院,英国伦敦NW44BT

出　　处：《纺织高校基础科学学报》2021年第2期108-116,共9页Basic Sciences Journal of Textile Universities

基　　金：国家自然科学基金(12001417);陕西省科技厅软科学项目(2019KRM072)。

摘　　要：针对樽海鞘算法(SSA)求解精度低,收敛速度慢,易陷入局部最优等缺陷,提出一种基于指数惯性权重和自适应t分布变异的改进樽海鞘算法。通过用最优位置替代个体位置改进了跟随者位置更新公式,提升了其寻优能力。将带有随机扰动项的指数递减惯性权重引入到改进后的跟随者位置更新公式中,平衡了算法全局搜索和局部搜索能力。在搜索过程中对每次更新后的位置以一定概率进行自适应t分布变异,避免其陷入局部最优。将改进算法与4种算法在8个不同维度、峰度的测试函数上进行了对比测试,并将其应用到2种工程设计问题中,结果表明:改进算法具有更好的全局和局部搜索能力,以及更高的寻优精度和更快的收敛速度,同时在求解实际问题时也表现出良好性能。In view of the shortcomings of the salp swarm algorithm such as low accuracy,slow convergence speed and being easy to fall into local optimum,an improved salp swarm algorithm based on exponential inertia weight and adaptive t-distribution mutation was proposed.The follower position updating formula was improved by replacing the individual position with the optimal position to enhance its optimization ability.The exponential decreasing inertia weight with random disturbance was introduced into the improved follower position updating formula to balance the global search and local search capabilities of the algorithm.In the search process,adaptive t-distribution mutation is carried out for each updated position according to a certain probability to avoid falling into local optimum.The improved algorithm is compared with four algorithms in eight different dimensions and kurtosis test functions,then it is applied to two engineering design problems.The results show that it has better global and local search capabilities,higher optimization accuracy and faster convergence speed,and also shows good performance in solving practical problems.

关 键 词：樽海鞘算法 指数递减惯性权重 自适应 t分布变异 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]