粒子群算法种群结构与种群多样性的关系研究

本文采用种群熵刻画粒子群算法中种群的多样性，并通过一组标准测试函数分析了不同的种群结构及邻城关系对粒子群算法种群多样性变化和算法性能的影响。实验结果表明，不同的种群结构具有不同的信息传递特性，对粒子群种群多样性的变化及算法的搜索效果有着较大的影响，因此对于不同的问题需根据其特点选择不同的种群结构。

Relations between Population Structure and Population Diversity of Particle Swarm Optimization Algorithm

This article adopts population entropy to portray the diversity of population, and has analyzed the influence of the different topology to the population entropy through a group of standard test functions. The test result indicates that, as a result of its own characteristic, different population structure has the tremendous influence to the population diversity and search effect of the algorithm, therefore when solving different problem, we should choose different popu- lation structure according to the characteristic of the problem.