求解连续空间优化问题的改进入侵杂草算法  

A Modified Invasive Weed Algorithm to Solving Continuous Space Optimization Problems

作　　者：朱丽娜[1] 李爽[2] ZHU Li-na1, LI Shuang2(1. Department of Basic, Zhengzhou Vocational College of Industrial Safety, Zhengzhou 455000, China 2. School of Mathematics and Information Science, Henan Normal University, Xinxiang 453007, Henan, Chin)

机构地区：[1]郑州工业安全职业学院基础部,河南郑州455000 [2]河南师范大学数学与信息科学学院,河南新乡453007

出　　处：《内蒙古师范大学学报（自然科学汉文版）》2018年第1期7-11,15,共6页Journal of Inner Mongolia Normal University(Natural Science Edition)

基　　金：河南省高等学校重点科研项目(18A110021)

摘　　要：针对标准入侵杂草算法在求解连续空间优化问题时存在易陷入局部最优、过早收敛的缺点,提出了一种改进入侵杂草算法.改进算法采用反向学习方法构建高质量的初始解,并综合考虑当前种群的适应度值和役龄水平确定各杂草的种子数目.同时,提出了基于分治思想的空间扩散方法以强化算法的搜索性能,并引入个体淘汰机制以避免早熟.最后,对五个测试函数进行了仿真,并与多种算法进行比较,结果表明改进算法的搜索性能得到了显著提升.The paper presents a modified metaheristic based on the invasive weed optimization（IWO）algorithm.The modified invasive weed optimization（MIWO）method aims to tackle the dilemma of easily trapping in local optimum and premature convergence when solving continuous space optimization problems.Firstly,an opposition-based learning technique is employed to generate initial solution with high quality.Meanwhile,the proposed MIWO determines the number of each seed considering both its fitness and age.A divide-and-conquer technique is introduced to enhance the exploring ability of original IWO.Moreover,MIWO adopts individual replacement mechanism to avoid premature convergence.Finally,the simulations were conducted on five benchmark functions,and the results are compared to those obtained by other metaheuristic.The experiment demonstrates that the modified algorithm has efficiently enhanced its exploring ability.

关 键 词：入侵杂草算法 反向学习 分治思想 连续空间优化问题 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]