种群规模可变的水波优化算法  

Water wave optimization with variable-size population

作　　者：张杰峰[1] 郑宇军[1] 

机构地区：[1]浙江工业大学计算机科学与技术学院,杭州310023

出　　处：《南京大学学报（自然科学版）》2015年第4期850-857,共8页Journal of Nanjing University（Natural Science）

基　　金：国家自然科学基金(61473263)

摘　　要：水波优化(water wave optimization,WWO)算法是一种基于浅水波理论的新兴智能优化算法,在大量基准问题上表现出了极为优越的性能.提出该算法的一个改进版本,将原算法中的固定种群规模改为线性递减的种群规模,从而在搜索的早期更好地支持全局探索,而在搜索的后期更多地进行局部开发.通过在IEEE CEC 2014测试集30个函数优化问题上的实验比较,改进后的算法在12个问题上显著优于原WWO算法,在4个问题上劣于原算法,在14个问题上无显著性差异.此结果表明提出的种群可变策略能够提高WWO算法的综合性能.Inspired by the shallow wave theory,water wave optimization（WWO）is a novel heuristic optimization algorithm which has shown significant performance advantage over a number of classical and recent evolutionary algorithms on a variety of benchmark problems.The WWO algorithm uses three types of operations：the propagation operator that makes high fitness waves search small areas and low fitness waves explore large areas,the refraction operator that helps waves to escape search stagnation and improves the diversity of the population,and the breaking operator that enables an intensive search around a potentially promising area.However,the original WWO uses a fixed population size which may limit the self-adaptivity to different states.To tackle this issue,in this paper we propose an improved version of WWO that uses a variable-size population instead of a fixed one.In more details,the new algorithm linearly decreases its population size with the progress of the search,using more individuals to facilitate global exploration in early stages while using fewer individuals to accelerate convergence in later stages,and thus achieving a better balance between global search and local search.The comparative experiments on 30 function optimization problems of the IEEE CEC 2014 test suite show that the improved algorithm is significantly better than the original WWO on 12 problems,worse than that on 4problems,and there is no significant difference between them on the remaining 14 problems.This demonstrates that the proposed variable population size strategy can effectively improve the overall performance of WWO.This is a first study on the variable population size in WWO.

关 键 词：进化算法 水波优化 种群规模 全局优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]