一种基于GEP的多层物流网络Prüfer编码优化算法  

Prüfer Code Optimization Algorithm for Multi-layer Logistic Networks Based on Gene Expression Programming

作　　者：李康顺[1] 黄浩华[1] 张文生[2] 

机构地区：[1]华南农业大学信息学院,广州510642 [2]中国科学院自动化研究所,北京100190

出　　处：《系统仿真学报》2012年第3期594-602,共9页Journal of System Simulation

基　　金：国家自然科学基金(70971043)

摘　　要：在对Prüfer编码研究的基础上,提出了一种基于基因表达式程序设计的多层物流网络Prüfer编码优化算法(GEPPCOA)。该算法一方面利用基因表达式程序设计的多基因结构特点,克服了传统演化算法中不能直接用个体表示多层物流模式的缺点;另一方面还对基因表达式程序设计原有的操作算子进行了针对Prüfer编码的改进。实验结果表明,使用GEPPCOA求解多层物流网络的Prüfer编码优化问题时,其各项支出费用性能指标均要好于传统的演化算法,提高了算法的收敛精度。A new Prüfer code optimization algorithm for multi-layer logistic networks based on gene expression programming（GEPPCOA） was proposed.On the one hand,in this algorithm,the shortcomings that the individuals in traditional evolutionary algorithm could not be used to represent the complex multi-layer logistic networks directly were overcome by the character of multi-gene structure in GEP.On the other hand,the evolution operators in GEP were improved according to the characters of Prüfer codes.The simulated experiments show that the performance＇s indices of the payments calculated by using GEPPCOA for optimizing the Prüfer codes of complex multi-layer logistic networks are better,which can get more precise solutions speed compared with the traditional evolution algorithms.

关 键 词：演化算法 基因表达式程序设计 Prüfer编码 多层物流网络 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]