Parameter Estimation Using Unscented Kalman Filter on the Gray-Box Model for Dynamic EEG System Modeling

Sun Hee Kim; Hyung Jeong Yang; Ngoc Anh Thi Nguyen; Raja Majid Mehmood; Seong Whan Lee

doi:10.1109/TIM.2020.2967138

Parameter Estimation Using Unscented Kalman Filter on the Gray-Box Model for Dynamic EEG System Modeling

Sun Hee Kim, Hyung Jeong Yang, Ngoc Anh Thi Nguyen, Raja Majid Mehmood, Seong Whan Lee

Department of Artificial Intelligence

Research output: Contribution to journal › Article › peer-review

9 Citations (Scopus)

Abstract

Model parameters' estimation is one of the most important tasks in the analysis and design process of a nonlinear dynamic system in real time, especially in the presence of noise. This article presents a novel approach in estimating important parameters of gray-box model for such a system on real nonlinear EEG to simulate efficiently the dynamic characteristics of neurons. Specifically, the proposed methodology exploits unscented Kalman filter (UKF) that is combined with chaos neural population model to formulate the interaction between the cortical areas. The proposed methodology is compared with the state-of-the-art parameters' estimation techniques to verify the efficiency of the UKF on the gray-box model. Experimental results show that the proposed method demonstrates the lowest error value of root mean square error (RMSE) among existing parameter estimation methods. The robustness of the proposed approach is further validated in its convergence and automation, with minimum error relatively than others and without any user-specified input, respectively.

Original language	English
Article number	8962161
Pages (from-to)	6175-6185
Number of pages	11
Journal	IEEE Transactions on Instrumentation and Measurement
Volume	69
Issue number	9
DOIs	https://doi.org/10.1109/TIM.2020.2967138
Publication status	Published - 2020 Sept

Bibliographical note

Funding Information:
Manuscript received August 27, 2019; revised December 15, 2019; accepted December 29, 2019. Date of publication January 17, 2020; date of current version August 11, 2020. This work was supported in part by the National Research Foundation of Korea (NRF) grant funded by the Korea Government (MSIT) under Grant NRF-2018R1A2B6006046, in part by a Korea University Grant, in part by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education under Grant NRF-2017R1A4A1015559, and in part by the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea Government under Grant 2017-0-00451 (Development of BCI Based Brain and Cognitive Computing Technology for Recognizing User’s Intentions Using Deep Learning). The Associate Editor coordinating the review process was Anirban Mukherjee. (Corresponding author: Seong-Whan Lee.) Sun-Hee Kim and Seong-Whan Lee are with the Department of Brain and Cognitive Engineering, Korea University, Seoul 02841, South Korea (e-mail: sunheekim@korea.ac.kr; sw.lee@korea.ac.kr).

Publisher Copyright:
© 1963-2012 IEEE.

Keywords

Electroencephalogram
gray-box model
nonlinear dynamic system
parameter estimation
unscented Kalman filter (UKF)

ASJC Scopus subject areas

Instrumentation
Electrical and Electronic Engineering

Access to Document

10.1109/TIM.2020.2967138

Cite this

@article{b2c74a38fc654728b981169103182182,

title = "Parameter Estimation Using Unscented Kalman Filter on the Gray-Box Model for Dynamic EEG System Modeling",

abstract = "Model parameters' estimation is one of the most important tasks in the analysis and design process of a nonlinear dynamic system in real time, especially in the presence of noise. This article presents a novel approach in estimating important parameters of gray-box model for such a system on real nonlinear EEG to simulate efficiently the dynamic characteristics of neurons. Specifically, the proposed methodology exploits unscented Kalman filter (UKF) that is combined with chaos neural population model to formulate the interaction between the cortical areas. The proposed methodology is compared with the state-of-the-art parameters' estimation techniques to verify the efficiency of the UKF on the gray-box model. Experimental results show that the proposed method demonstrates the lowest error value of root mean square error (RMSE) among existing parameter estimation methods. The robustness of the proposed approach is further validated in its convergence and automation, with minimum error relatively than others and without any user-specified input, respectively.",

keywords = "Electroencephalogram, gray-box model, nonlinear dynamic system, parameter estimation, unscented Kalman filter (UKF)",

author = "Kim, {Sun Hee} and Yang, {Hyung Jeong} and Nguyen, {Ngoc Anh Thi} and Mehmood, {Raja Majid} and Lee, {Seong Whan}",

note = "Funding Information: Manuscript received August 27, 2019; revised December 15, 2019; accepted December 29, 2019. Date of publication January 17, 2020; date of current version August 11, 2020. This work was supported in part by the National Research Foundation of Korea (NRF) grant funded by the Korea Government (MSIT) under Grant NRF-2018R1A2B6006046, in part by a Korea University Grant, in part by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education under Grant NRF-2017R1A4A1015559, and in part by the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea Government under Grant 2017-0-00451 (Development of BCI Based Brain and Cognitive Computing Technology for Recognizing User{\textquoteright}s Intentions Using Deep Learning). The Associate Editor coordinating the review process was Anirban Mukherjee. (Corresponding author: Seong-Whan Lee.) Sun-Hee Kim and Seong-Whan Lee are with the Department of Brain and Cognitive Engineering, Korea University, Seoul 02841, South Korea (e-mail: sunheekim@korea.ac.kr; sw.lee@korea.ac.kr). Publisher Copyright: {\textcopyright} 1963-2012 IEEE.",

year = "2020",

month = sep,

doi = "10.1109/TIM.2020.2967138",

language = "English",

volume = "69",

pages = "6175--6185",

journal = "IEEE Transactions on Instrumentation and Measurement",

issn = "0018-9456",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "9",

}

TY - JOUR

T1 - Parameter Estimation Using Unscented Kalman Filter on the Gray-Box Model for Dynamic EEG System Modeling

AU - Kim, Sun Hee

AU - Yang, Hyung Jeong

AU - Nguyen, Ngoc Anh Thi

AU - Mehmood, Raja Majid

AU - Lee, Seong Whan

N1 - Funding Information: Manuscript received August 27, 2019; revised December 15, 2019; accepted December 29, 2019. Date of publication January 17, 2020; date of current version August 11, 2020. This work was supported in part by the National Research Foundation of Korea (NRF) grant funded by the Korea Government (MSIT) under Grant NRF-2018R1A2B6006046, in part by a Korea University Grant, in part by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education under Grant NRF-2017R1A4A1015559, and in part by the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea Government under Grant 2017-0-00451 (Development of BCI Based Brain and Cognitive Computing Technology for Recognizing User’s Intentions Using Deep Learning). The Associate Editor coordinating the review process was Anirban Mukherjee. (Corresponding author: Seong-Whan Lee.) Sun-Hee Kim and Seong-Whan Lee are with the Department of Brain and Cognitive Engineering, Korea University, Seoul 02841, South Korea (e-mail: sunheekim@korea.ac.kr; sw.lee@korea.ac.kr). Publisher Copyright: © 1963-2012 IEEE.

PY - 2020/9

Y1 - 2020/9

N2 - Model parameters' estimation is one of the most important tasks in the analysis and design process of a nonlinear dynamic system in real time, especially in the presence of noise. This article presents a novel approach in estimating important parameters of gray-box model for such a system on real nonlinear EEG to simulate efficiently the dynamic characteristics of neurons. Specifically, the proposed methodology exploits unscented Kalman filter (UKF) that is combined with chaos neural population model to formulate the interaction between the cortical areas. The proposed methodology is compared with the state-of-the-art parameters' estimation techniques to verify the efficiency of the UKF on the gray-box model. Experimental results show that the proposed method demonstrates the lowest error value of root mean square error (RMSE) among existing parameter estimation methods. The robustness of the proposed approach is further validated in its convergence and automation, with minimum error relatively than others and without any user-specified input, respectively.

AB - Model parameters' estimation is one of the most important tasks in the analysis and design process of a nonlinear dynamic system in real time, especially in the presence of noise. This article presents a novel approach in estimating important parameters of gray-box model for such a system on real nonlinear EEG to simulate efficiently the dynamic characteristics of neurons. Specifically, the proposed methodology exploits unscented Kalman filter (UKF) that is combined with chaos neural population model to formulate the interaction between the cortical areas. The proposed methodology is compared with the state-of-the-art parameters' estimation techniques to verify the efficiency of the UKF on the gray-box model. Experimental results show that the proposed method demonstrates the lowest error value of root mean square error (RMSE) among existing parameter estimation methods. The robustness of the proposed approach is further validated in its convergence and automation, with minimum error relatively than others and without any user-specified input, respectively.

KW - Electroencephalogram

KW - gray-box model

KW - nonlinear dynamic system

KW - parameter estimation

KW - unscented Kalman filter (UKF)

UR - http://www.scopus.com/inward/record.url?scp=85089979924&partnerID=8YFLogxK

U2 - 10.1109/TIM.2020.2967138

DO - 10.1109/TIM.2020.2967138

M3 - Article

AN - SCOPUS:85089979924

SN - 0018-9456

VL - 69

SP - 6175

EP - 6185

JO - IEEE Transactions on Instrumentation and Measurement

JF - IEEE Transactions on Instrumentation and Measurement

IS - 9

M1 - 8962161

ER -

Parameter Estimation Using Unscented Kalman Filter on the Gray-Box Model for Dynamic EEG System Modeling

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this