CIESC Journal ›› 2017, Vol. 68 ›› Issue (8): 3161-3167.DOI: 10.11949/j.issn.0438-1157.20161786

Previous Articles     Next Articles

Biogeography-based learning particle swarm optimization method for solving dynamic optimization problems in chemical processes

CHEN Xu1,2, MEI Congli1, XU Bin3, DING Yuhan1, LIU Guohai1   

  1. 1 School of Electrical and Information Engineering, Jiangsu University, Zhenjiang 212013, Jiangsu, China;
    2 Key Laboratory of Advanced Control and Optimization for Chemical Processes, Ministry of Education, East China University of Science and Technology, Shanghai 200237, China;
    3 School of Mechanical Engineering, Shanghai University of Engineering Science, Shanghai 201620, China
  • Received:2016-12-30 Revised:2017-03-19 Online:2017-08-05 Published:2017-08-05
  • Supported by:

    supported by Natural Science Foundation of Jiangsu Province (BK20160540, BK20130531), the Research Talents Startup Foundation of Jiangsu University (15JDG139), China Postdoctoral Science Foundation (2016M591783) and the Fundamental Research Funds for the Central Universities (222201717006).

一种求解过程动态优化问题的生物地理学习粒子群算法

陈旭1,2, 梅从立1, 徐斌3, 丁煜函1, 刘国海1   

  1. 1 江苏大学电气信息工程学院, 江苏 镇江 212013;
    2 华东理工大学化工过程先进控制和优化技术教育部重点实验室, 上海 200237;
    3 上海工程技术大学机械工程学院, 上海 201620
  • 通讯作者: 陈旭
  • 基金资助:

    江苏省自然科学基金项目(BK20160540,BK20130531);江苏大学人才启动基金项目(15JDG139);中国博士后科学基金项目(2016M591783);中央高校基本科研业务费重点科研基地创新基金项目(222201717006)。

Abstract:

Intelligent optimization algorithms have been playing an increasing role in dynamic optimization, due to advantages of wide applicability and strong global searching capability. Biogeography-based learning particle swarm optimization (BLPSO) was proposed for dynamic optimization problems (DOPs) by hybridizing biogeography-based and particle swarm optimization. BLPSO employed a new biogeography-based learning approach for construction of learning examples by ranking of particles (i.e., the quality of particles) and dimension as unit, such that learning efficiency was enhanced. Control vector parameterization first converted DOPs into nonlinear programming problems which were then solved by BLPSO. The simulation results on typical DOPs with non-differentiable, multi-modal and multi-variable characteristics show that BLPSO has outstanding solution precision and convergence speed.

Key words: global optimization, dynamics, algorithm, control vector parameterization, biogeography-based learning particle swarm optimization

摘要:

智能优化算法具有适用性广泛、全局搜索能力强等优点,近年来在动态优化中的应用逐渐增多。通过混合生物地理优化与粒子群优化,提出了生物地理学习粒子群(biogeography-based learning particle swarm optimization,BLPSO)算法,并用于动态优化问题的求解。BLPSO采用了新型的生物地理学习方式,该方式根据粒子“排名”,即粒子的优劣,以维度为单位构造学习粒子,提高了学习的效率。针对动态优化问题,首先通过控制向量参数化将其转化为非线性规划问题,然后采用BLPSO算法进行求解。最后,将BLPSO应用于非可微、多峰、多变量等典型动态优化问题的求解,计算结果表明BLPSO具有较好的搜索精度和收敛速度。

关键词: 全局优化, 动态学, 算法, 控制向量参数化, 生物地理学习粒子群算法

CLC Number: