正交Alopex算法在机械优化设计中的应用  

Orthogonal Alopex algorithm and its application on mechanical optimization design

作　　者：张延萍[1] 杜新峰[1] 卢青波[1,2] 

机构地区：[1]郑州职业技术学院,郑州450121 [2]太原科技大学机电工程学院,太原030024

出　　处：《现代制造工程》2012年第7期70-73,共4页Modern Manufacturing Engineering

基　　金：河南省科技攻关计划项目(2010C46008);河南省高等学校青年骨干教师资助计划项目(2011GGJS_296)

摘　　要：针对Alopex算法对初值敏感的特性及正交设计可以使初始种群在空间分布更加均匀的特点,提出了正交Alopex算法。该算法为了克服后期"早熟"收敛的缺陷,引入了变异操作。数值仿真结果及两个工程优化问题的求解结果表明,所提出的算法能有效避免"早熟"收敛,且求解结果优于传统优化及比较算法。The performance of the Alopex algorithm depends on the initial population distribution. The orthogonal design can gen- erate more uniform distribution initial population. An orthogonal Alopex algorithm is presented based on the above two points. In order to overcome the premature convergence, the new algorithm incorporated the mutation operation. The results of numerical ex- perimental and two engineering optimization design examples show that the proposed algorithm can overcome the premature con- vergence efficiently, and the results are better than the traditional optimization algorithm and the comparison algorithms.

关 键 词：ALOPEX算法 变异 早熟 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]