论文部分内容阅读
提出一种自适应动态重组粒子群优化算法.该算法采用凝聚的层次聚类算法,将种群分成若干个子群体,用一个精英集对非支配解进行存储;根据贡献度和多样性,对各子群体的粒子和整个种群进行自适应动态重组;同时引入扰动算子对精英集存储的非支配解进行扰动,实现对精英集进行动态调整.利用具有不同特点的测试函数进行验证并与同类算法相比较,结果表明,所提出的算法可加快收敛速度,提高种群的可进化能力.
An adaptive dynamic reconfigurable particle swarm optimization algorithm is proposed.The algorithm uses a cohesive hierarchical clustering algorithm that divides the population into several subgroups and stores the nondominated solution with an elitist set.According to the contributions and diversity, Population particles and the entire population are adaptively and dynamically reorganized. At the same time, perturbation operators are introduced to perturb the nondominated solution of the elitist set storage, and the dynamic adjustment of the elitist sets is achieved. The test functions with different characteristics are used to verify and compare with the same algorithm The results show that the proposed algorithm can speed up the convergence and improve the evolvability of the population.