粒子群复形法求解生物反应器的补料优化  

Complex Particle Swarm Optimization Algorithm for Solving Feed Policies Optimization for Bioreactor

作　　者：刘贺同[1] 莫愿斌[1] 

机构地区：[1]中国计量学院理学院,浙江杭州310018

出　　处：《机床与液压》2007年第5期94-97,共4页Machine Tool & Hydraulics

摘　　要：动态系统优化是过程设计、操作及控制的一个重要课题,但有效地求出其全局最优解是困难的。现今各种算法求解此问题时需要增强算法的优化性能和算法的简洁性;而遗传算法求此类问题时需要进行离散化,本文为此提出用粒子群复形法求解此类问题。此算法是在充分考虑粒子群算法与复形法的特性,保持复形法的迭代机理,同时加入粒子群算法基本思想,再结合几何分点、梯度与信赖域等的基础上提出的;以克服粒子群算法与复形法易陷局部极值的不足。性能测试的结果表明:该算法简便、可行、高效;最后将所提出的算法用于Park-Ram irez生物反应器补料流率的动态优化,也取得了满意的效果。Dynamic system optimization is an important problem in process design, manipulation and control, but it is difficult to effectively find out the global optimization solution. The existed problems of the algorithm used now are that the algorithm is easy to be trapped into local minima in optimizing model and the manipulation of the algorithm is not convenienL Aiming at these problems, a new algorithm, complex particle swarm optimization algorithm was developed, which combines the advantages of method of complex （MC）, particle swarm optimization （PSO）, geometric point, gradient and the confidence region method and overcomes the disadvantage that PSO and MC are easy to be trapped into local minima in optimizing multimodal function. The result of a case study shows that CPSO is convenient, feasible, and efficient. When applied to dynamic optimizing feed-rate of Park-Ramirez bioreactor, satisfying results were gotten.

关 键 词：动态优化 复形法 粒子群算法 Park—Ramirez 生物反应器 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]