Journal of Computer Applications ›› 2015, Vol. 35 ›› Issue (9): 2584-2589.DOI: 10.11772/j.issn.1001-9081.2015.09.2584

Previous Articles     Next Articles

Chromosomal translocation-based Dynamic evolutionary algorithm

TAN Yang1,2, NING Ke2, CHEN Lin2   

  1. 1. College of Mathematics and Computer Science, Hunan Normal University, Changsha Hunan 410081, China;
    2. Department of Information Engineering, Hunan Radio and Television University, Changsha Hunan 410004, China
  • Received:2015-04-15 Revised:2015-06-27 Online:2015-09-10 Published:2015-09-17


谭阳1,2, 宁可2, 陈琳2   

  1. 1. 湖南师范大学 数学与计算机科学学院, 长沙 410081;
    2. 湖南广播电视大学 信息工程系, 长沙 410004
  • 通讯作者: 谭阳(1979-),男,湖南望城人,副教授,硕士,主要研究方向:人工智能、信息安全、数据挖掘,
  • 作者简介:宁可(1983-),男,湖南武冈人,讲师,硕士,主要研究方向:人工智能;陈琳(1980-),女,湖南湘潭人,讲师,硕士,主要研究方向:数据挖掘。
  • 基金资助:

Abstract: When traditional binary-coded evolutionary algorithms are applied to optimize functions, the mutual interference between different dimensions would prevent effective restructuring of some low-order modes. A new evolutionary algorithm, called Dynamic Chromosomal Translocation-based Evolutionary Algorithm (CTDEA), was proposed based on cytological findings. This algorithm simulated the structuralized process of organic chromosome inside cells by constructing gene matrixes, and realized modular translocations of homogeneous chromosomes on the basis of gene matrix, in order to maintain the diversity of populations. Moreover, the individual fitness-based population-dividing method was adopted to safeguard elite populations, ensure competitions among individuals and improve the optimization speed of the algorithm. Experimental results show that compared with existing Genetic Algorithm (GA) and distribution estimation algorithms, this evolutionary algorithm greatly improves the population diversity, keeping the diversity of populations around 0.25. In addition, this algorithm shows obvious advantages in accuracy, stability and speed of optimization.

Key words: chromosomal translocation, evolutionary algorithm, gene matrix, modularization, function optimization

摘要: 针对采用二进制编码的进化算法在函数优化过程中会因为维度之间的相互干扰,导致部分低阶模式出现无法进行有效重组的现象,提出一种新的结合细胞学研究成果的进化算法——染色体易位的动态进化算法(CTDEA)。算法通过构建基因矩阵来模拟有机染色体在细胞内的结构化过程,并在基因矩阵的基础上对出现同质化的染色体短列实施模块化的易位操作,以此来维护种群的多样性;同时通过个体适应度划分种群的方式来维护精英个体,确保个体间的竞争压力,提升算法的寻优速度。实验结果表明,该进化算法与已有的遗传算法(GA)和分布估计算法相比较,在维护种群多样性方面有较大改进,能够将种群的多样性保持在0.25左右;且在寻优的精度、稳定性以及速度上也有明显的改进和提高。

关键词: 染色体易位, 进化算法, 基因矩阵, 模块化, 函数优化

CLC Number: