引用本文:周博,严洪森.一类具有羊群效应的多重多维泰勒网动力学模型[J].控制理论与应用,2015,32(7):963~969.[点击复制]
ZHOU Bo,YAN Hong-sen.A class of multiple multi-dimensional Taylor networks dynamics model with herd behavior[J].Control Theory and Technology,2015,32(7):963~969.[点击复制]
一类具有羊群效应的多重多维泰勒网动力学模型
A class of multiple multi-dimensional Taylor networks dynamics model with herd behavior
摘要点击 2564  全文点击 899  投稿时间：2014-07-04  修订日期：2015-01-21
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2015.40628
  2015,32(7):963-969
中文关键词  多重多维泰勒网  间歇反馈  羊群效应  动力学模型  系统辨识
英文关键词  multiple multi-dimensional Taylor network  intermittent feedback  herd behavior  dynamics model  system identification
基金项目  国家自然科学基金项目(50875046, 60934008), 中央高校基本科研业务费专项资金项目(2242014K10031)资助.
作者单位E-mail
周博* 东南大学 自动化学院 复杂工程系统测量与控制教育部重点实验室 ai_haibara@126.com 
严洪森 东南大学 自动化学院 复杂工程系统测量与控制教育部重点实验室  
中文摘要
      含大量有主观判断力个体参与的系统, 往往会表现出羊群效应的特点, 现有研究方法主要从机理和个体角度展开, 不适合建立整体模型. 本文提出带间歇反馈的多重多维泰勒网动力学模型建模方法, 适合建立含有羊群效应系统的整体优化模型. 首先讨论了多维泰勒网和间歇反馈模型建立方法. 然后研究了多重多维泰勒网的特性, 并通过多重多维泰勒网调节羊群效应和系统长期趋势两个部分在系统中合理比重. 最后给出了辨识模型参数的具体方法和步骤. 应用实例的结果验证了带间歇反馈的多重多维泰勒网建模方法在实际应用中的可实现性, 同时其拥有更好的预测精度.
英文摘要
      A system that involves considerable quantities of individuals with subjective judgment tends to exhibit the feature of herd behavior. Existing research methods mainly focus on mechanisms and individuals, and hence they are incapable of establishing integrated models. We propose a dynamics modeling method for multiple multi-dimensional Taylor networks with intermittent feedback. The method is proved to be effective in establishing an overall optimization model of a system with herd behavior. Firstly, the methods of modeling multi-dimensional Taylor networks with intermittent feedback are discussed. Secondly, the characteristic of multiple multi-dimensional Taylor network is studied and used to regulate the appropriate proportion of herding behavior to long-term trend in the system. Finally, the specific method and procedure of identifying model parameters are given. The result of the application example demonstrates the modeling method of multiple multi-dimensional Taylor networks with intermittent feedback is realizable in practical applications, and has higher prediction accuracy.