基于Alopex的野草算法  

Improved invasive weed optimization algorithm based on Alopex

作　　者：彭俊[1] 詹泳[1] 

机构地区：[1]华南师范大学计算机学院,广东广州510631

出　　处：《计算机工程与设计》2014年第12期4202-4206,共5页Computer Engineering and Design

摘　　要：针对野草算法存在求解精度不高、收敛速度慢的问题,提出一种基于Alopex的野草算法。在原有野草算法框架的空间扩散阶段引入Alopex算法,通过从父代和子代个体自变量和目标函数值的变化情况获得启发信息,指导种群向最优方向进化。结合后的算法能够充分发挥两者的优点,改善野草算法收敛速度以及寻优精度。对典型基准函数的测试结果表明,该算法要优于基本野草算法,表现更为稳定,体现出较好的全局搜索能力,具有更快的收敛速度和更高的寻优精度,更适合于解决其它算法难以解决的高维多峰值函数的优化问题;通过与其它相关智能算法的比较,进一步验证了该算法的有效性。Concerning the low accuracy in application and the slow rate of convergence of invasive weed optimization,a new invasive weed optimization（IWO）based on the algorithm of pattern extraction（Alopex）was proposed.The Alopex was imported in the spatial dispersal of IWO’s base framework.The inspired information was obtained from changes of individuals making the population evolve in the direction of the global optimal.Based on the novel and distinct qualifications of IWO and Alopex,the AIWO algorithm was introduced and their excellent features in this extended algorithm were combined.The efficiency of this algorithm in the speed of convergence and the optimality of results were compared with IWO through a number of common benchmark functions.The results of experiment reveal that the proposed algorithm is more stable than IWO and shows better capability of global search,which makes it suitable for the optimization of multi-dimensional functions of which some other algorithms are incapable.Besides the experiment results,comparing with some other related intelligent algorithms also give further confirmation of the effectiveness of the proposed algorithm.

关 键 词：进化算法 野草算法 模式提取 数值优化 全局最优 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]