计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (10): 7-10.DOI: 10.3778/j.issn.1002-8331.2009.10.003

• 博士论坛 • 上一篇    下一篇

非线性动态系统建模方法研究

王 峰1,2,邢科义1,徐小平3   

  1. 1.西安交通大学 系统工程研究所 机械制造系统工程国家重点实验室,西安 710049
    2.西安交通大学 理学院,西安 710049
    3.西安理工大学 理学院,西安 710048
  • 收稿日期:2008-12-04 修回日期:2009-01-07 出版日期:2009-04-01 发布日期:2009-04-01
  • 通讯作者: 王 峰

Study on modeling method of nonlinear dynamic system

WANG Feng1,2,XING Ke-yi1,XU Xiao-ping3   

  1. 1.State Key Lab for Manufacturing System Eng.,System Eng. Institute,Xi’an Jiaotong University,Xi’an 710049,China
    2.School of Sciences,Xi’an Jiaotong University,Xi’an 710049,China
    3.School of Sciences,Xi’an University of Technology,Xi’an 710048,China
  • Received:2008-12-04 Revised:2009-01-07 Online:2009-04-01 Published:2009-04-01
  • Contact: WANG Feng

摘要: 讨论了一类非线性动态系统建模的新方法。首先,假设原非线性动态系统可以用Hammerstein模型来表示。然后,将Hammerstein模型的非线性传递函数转换为等价的线性形式,从而建立起中间模型。接下来,利用粒子群优化(Particle Swarm Optimization,PSO)算法辨识出中间模型参数。最后,通过中间模型参数与Hammerstein模型参数之间的关系,推出非线性静态环节和线性动态环节的参数,从而实现原非线性动态系统建模。为了进一步增强建模的性能,提出了利用一种改进的粒子群优化(Improved Particle Swarm Optimization,IPSO)算法。仿真结果说明了该方法的合理性和有效性。

关键词: 非线性动态系统, 建模, Hammerstein模型, 粒子群优化(PSO)算法

Abstract: A novel modeling method for a type of nonlinear dynamic system is discussed in this paper.First of all,it is supposed that the original nonlinear dynamic system can be expressed by Hammerstein model. And then,the nonlinear transfer function of Hammerstein model is converted to the same form as linear one,thus generating the intermediate model.After that,the parameters of the intermediate model are obtained using Particle Swarm Optimization(PSO) algorithm.Finally,through the relations of the parameters of intermediate model and that of Hammerstein model,the parameters of the nonlinear static subunit and linear dynamic subunit are derived.Thus the original nonlinear dynamic system is modeled.In order to further enhance the performance of modeling,an Improved Particle Swarm Optimization(IPSO) algorithm is also proposed.The rationality and efficiency of the presented algorithm are demonstrated by simulation examples.

Key words: nonlinear dynamic system, modeling, Hammerstein model, Particle Swarm Optimization(PSO) algorithm