基于高斯分布和模拟退火算法的免疫微粒群优化算法研究

Research of immune particle swarm optimization algorithm based on Gaussian distribution and simulated annealing algorithm

针对微粒群算法在搜索过程中粒子容易失去多样性而陷入局部最优且搜索速度较慢的缺陷,提出了一种基于高斯分布和模拟退火算法的免疫微粒群算法,该算法借助高斯分布和模拟退火的有关机理,分别进行免疫接种和免疫选择的操作。使用常用的基准函数对算法进行了仿真验证工作,通过与全局微粒群优化算法、变惯性权值微粒群优化算法的对比表明,免疫微粒群优化算法(IPSO)在搜索速度和全局寻优方面具有一定的优势。

The particle swarm optimization algorithm is not only easy to lose diversity and run into local optimization in course of search, but also the speed of search is low. This article presented an inmmune Particle Swarm Optimization （PSO） algorithm through inmmune inoculation and immune choice, which recurs to mechanism of GUASS and SA. The common norm function is used to develop simulated and validated work, Comparison of simulated results between Inmmune Particle Swarm Optimization （IPSO）, PSO and DWIPSO shows that IPSO has the advantage of improving the global search ability and decreasing calculated steps.