进化算法的新界定  

New Definition of Evolutionary Algorithm

作　　者：王怀晓[1,2] 刘建永[1] 印祝宸[1,3] 付成群[1] 

机构地区：[1]解放军理工大学野战工程学院,南京210007 [2]蚌埠汽车士官学校,蚌埠233011 [3]南京炮兵学院,南京211132

出　　处：《计算机与数字工程》2015年第8期1387-1389,1542,共4页Computer & Digital Engineering

摘　　要：传统的进化算法在二十世纪六十年代提出,只包括遗传算法、遗传规划、进化策略和进化规划四种方法,随后出现的进化机制的方法没有包含在进化算法的分类之中,这一状况导致了对进化算法研究的局限性,因此需要对进化算法进行重新界定。把新产生的具有进化机制的算法纳入进化算法的范围之中,并使进化算法的内涵具有发展性,提出了新的进化算法的分类方法,新的进化算法的分类方法更加地科学、涵盖更广、有发展性和时效性。The traditional evolutionary algorithm is proposed in the 1960s, which only includes four methods that are genetic algorithm, genetic programming, evolution strategies and evolutionary programming. Following algorithms with evo- lutionary mechanism are not included in the classification of the evolutionary algorithm which leads to the limitations of re- search of evolutionary algorithm, so it is necessary to redefine evolutionary algorithm. New algorithm is incorporated with evolutionary mechanism into the scope of evolutionary algorithm and make the connotation of evolutionary algorithm can be developed. The new classification of evolutionary algorithm is proposed to make the classification of evolutionary algorithm more scientific, wider in the scope, with property of development and effectiveness for period of time.

关 键 词：进化算法 分类 智能计算 新界定 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]