文献类型：期刊文献

中文题名：两群微粒群优化算法及其应用

英文题名：Two sub-swarms particle swarm optimization algorithm and its application

作者：陈国初[1,2];俞金寿[2]

机构：[1]上海电机学院电气学院;[2]华东理工大学,自动化研究所,上海,200237

年份：2007

卷号：24

期号：2

起止页码：294

中文期刊名：控制理论与应用

外文期刊名：Control Theory & Applications

收录：CSTPCD;;EI(收录号：20072310642780);Scopus;北大核心:【北大核心2004】；CSCD:【CSCD2011_2012】；

基金：上海市教委自然科学科研项目(05vz01)

语种：中文

中文关键词：微粒群优化算法;优化;催化裂化装置:轻柴油95%点;软测量

外文关键词：PSO（particle swarm optimization algorithm）; optimization; fluid catalytic cracking unit; light diesel oil; soft-sensor

摘要：针对微粒群优化算法容易陷入局部极值的缺陷,提出两群微粒群优化算法．通过对5种常用测试函数进行测试和比较,结果表明两群微粒群优化算法比基本微粒群优化算法更容易找到全局最优解,优化效率明显提高．然后将两群微粒群优化算法用于催化裂化装置主分馏塔轻柴油95%点软测量建模,通过与实际工业数据对比,表明该软测量模型具有高的精度、好的性能和广阔的应用前景．
In order to improve optimization performance of particle swarm optimization algorithm（PSO）, a new two subswarms particle swarm optimization algorithm（TSPSO） is proposed in this paper. Then, both TSPSO and PSO are used to resolve five well-known and widely used test functions＇ optimization problems. Results show that TSPSO has greater efficiency and better performance than PSO. TSPSO is also applied to train artificial neural network（NN）to construct a practical soft-sensor for the 95%-point light diesel oil in a main fractionator of fluid catalytic cracking unit（FCCU）. The obtained results and comparison with actual industrial data indicate that the proposed method is feasible and effective in soft-sensor for the 95%-point light diesel oil.