基于万有引力思想的遗传算子  

Genetic Operator Based on the Idea of Universal Gravitation

作　　者：李松芳 刘伟[1] 

机构地区：[1]广东工业大学应用数学学院,广东广州510006

出　　处：《广东工业大学学报》2015年第1期121-127,共7页Journal of Guangdong University of Technology

基　　金：国家自然科学基金资助项目(60974077)

摘　　要：借鉴万有引力搜索算法和局部搜索思想,对遗传算法中的算术交叉和非均匀变异算子进行了改进,将万有引力搜索和混沌搜索引入新算子.新算子能够充分利用最优个体信息,兼具有全局搜索与局部搜索性能.仿真实验表明,新算子不但加快了算法的收敛速度,而且提高了求解精度.With reference to the thought of Gravity search and Local search,it makes improvement to the arithmetic crossover and non-uniform mutation. Gravity search and Chaotic search were integrated into the improved crossover and mutation operators. The new operators have global search and local search performance. At the same time,the best individual was used for chaotic search. Simulation experiments show that the improved operators not only speed up the convergence of the algorithm,but also improve the precision.

关 键 词：遗传算法 万有引力搜索 混沌 交叉算子 变异算子 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]