A method of parameter identification for dynamic systems based on model output minimum entropy

T.Y. Liu, J.F. Jia, H. Wang, H. Yue

Research output: Contribution to conferencePaper

Abstract

Parameter estimation is important in mathematical modeling. The Maximum Likelihood method can be used when the probability density function of observation is known. However, this assumption may not be satisfied in practice. To deal with this problem, a new parameter estimation method for dynamic systems is proposed using the entropy of probability density function for system output viable and two performance functions are also given. To illustrate the effectiveness of this method, HIV/AIDS model is taken as an example to evaluate simulation and results are encouraging.
Original languageEnglish
Pages112-114
Number of pages2
DOIs
Publication statusPublished - 2007
Event26th Chinese Control Conference - Zhangjiajie, China
Duration: 26 Jul 200731 Jul 2007

Conference

Conference26th Chinese Control Conference
CountryChina
CityZhangjiajie
Period26/07/0731/07/07

Fingerprint

Parameter estimation
Probability density function
Identification (control systems)
Dynamical systems
Entropy
Maximum likelihood

Keywords

  • dynamic systems
  • histogram
  • probability density function
  • entropy
  • parameter estimation

Cite this

Liu, T. Y., Jia, J. F., Wang, H., & Yue, H. (2007). A method of parameter identification for dynamic systems based on model output minimum entropy. 112-114. Paper presented at 26th Chinese Control Conference , Zhangjiajie, China. https://doi.org/10.1109/CHICC.2006.4346781
Liu, T.Y. ; Jia, J.F. ; Wang, H. ; Yue, H. / A method of parameter identification for dynamic systems based on model output minimum entropy. Paper presented at 26th Chinese Control Conference , Zhangjiajie, China.2 p.
@conference{e60ff53426424485a8c6d883d1069941,
title = "A method of parameter identification for dynamic systems based on model output minimum entropy",
abstract = "Parameter estimation is important in mathematical modeling. The Maximum Likelihood method can be used when the probability density function of observation is known. However, this assumption may not be satisfied in practice. To deal with this problem, a new parameter estimation method for dynamic systems is proposed using the entropy of probability density function for system output viable and two performance functions are also given. To illustrate the effectiveness of this method, HIV/AIDS model is taken as an example to evaluate simulation and results are encouraging.",
keywords = "dynamic systems, histogram , probability density function, entropy, parameter estimation",
author = "T.Y. Liu and J.F. Jia and H. Wang and H. Yue",
year = "2007",
doi = "10.1109/CHICC.2006.4346781",
language = "English",
pages = "112--114",
note = "26th Chinese Control Conference ; Conference date: 26-07-2007 Through 31-07-2007",

}

Liu, TY, Jia, JF, Wang, H & Yue, H 2007, 'A method of parameter identification for dynamic systems based on model output minimum entropy', Paper presented at 26th Chinese Control Conference , Zhangjiajie, China, 26/07/07 - 31/07/07 pp. 112-114. https://doi.org/10.1109/CHICC.2006.4346781

A method of parameter identification for dynamic systems based on model output minimum entropy. / Liu, T.Y.; Jia, J.F.; Wang, H.; Yue, H.

2007. 112-114 Paper presented at 26th Chinese Control Conference , Zhangjiajie, China.

Research output: Contribution to conferencePaper

TY - CONF

T1 - A method of parameter identification for dynamic systems based on model output minimum entropy

AU - Liu, T.Y.

AU - Jia, J.F.

AU - Wang, H.

AU - Yue, H.

PY - 2007

Y1 - 2007

N2 - Parameter estimation is important in mathematical modeling. The Maximum Likelihood method can be used when the probability density function of observation is known. However, this assumption may not be satisfied in practice. To deal with this problem, a new parameter estimation method for dynamic systems is proposed using the entropy of probability density function for system output viable and two performance functions are also given. To illustrate the effectiveness of this method, HIV/AIDS model is taken as an example to evaluate simulation and results are encouraging.

AB - Parameter estimation is important in mathematical modeling. The Maximum Likelihood method can be used when the probability density function of observation is known. However, this assumption may not be satisfied in practice. To deal with this problem, a new parameter estimation method for dynamic systems is proposed using the entropy of probability density function for system output viable and two performance functions are also given. To illustrate the effectiveness of this method, HIV/AIDS model is taken as an example to evaluate simulation and results are encouraging.

KW - dynamic systems

KW - histogram

KW - probability density function

KW - entropy

KW - parameter estimation

UR - http://ccc.amss.ac.cn/main/index.html

U2 - 10.1109/CHICC.2006.4346781

DO - 10.1109/CHICC.2006.4346781

M3 - Paper

SP - 112

EP - 114

ER -

Liu TY, Jia JF, Wang H, Yue H. A method of parameter identification for dynamic systems based on model output minimum entropy. 2007. Paper presented at 26th Chinese Control Conference , Zhangjiajie, China. https://doi.org/10.1109/CHICC.2006.4346781