Journal of the Korean Institute of Intelligent Systems (한국지능시스템학회논문지)

Korean Institute of Intelligent Systems (한국지능시스템학회)
권호 표지

Nonlinear System Modeling using Independent Component Analysis and Neuro-Fuzzy Method

독립 성분 분석기법과 뉴로-퍼지를 이용한 비선형 시스템 모델링

  • 김성수 (우석대학교 전기공학과) ;
  • 곽근창 (충북대학교 전기전자공학부) ;
  • 유정웅 (충북대학교 전기전자공학부)
  • Published : 2000.10.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, an efficient fuzzy rule generation scheme for adaptive neuro-fuzzy system modeling using the Independent Component Analysis(ICA) as a preprocessing is proposed. Correlation between inputs was not considered in the conventional neuro- fuzzy modeling schemes, such that enormous number of rules and large amount of error were unavoidable. The correlation between inputs is weakened by employing ICA so that the number of rules and the amount of error are reduced. In simulation, the Box-Jenkins furnace data is used to verify the effectiveness of the proposed method.

본 논문에서는 적응 뉴로-퍼지 모델링을 위해 최근에 BBS(blind source separation)분야에서 발전된 독립 성분 분석기법(ICA)을 전처리로 이용하여 효과적인 퍼지 규칙을 생성하는 방법을 제안한다. 기존의 뉴로-퍼지 모델링은 입력 데이터 성분간의 상관관계를 고려하지 않고 입력공간을 분할하기 때문에 효과적으로 분할하지 못하는 단점이 있다. 이로 인해 과도한 규칙 수와 큰 오차를 가지고 있었다. 이에, 본 연구에서는 독립 성분 분석기법을 이용하여 입력 데이터 성분간의 상관관계를 제거함으로서 적은 규칙 수를 갖으면서도 효율적인 퍼지 규칙을 얻을 수 있도록 하였다. 시뮬레이션 예로서 Box-Jenkins의 가스로 데이터의 모델링에 적용하여 유용성과 제안된 방법이 이전의 연구보다 좋은 결과를 보임을 알 수 있었다.

Keywords

References

  1. Information Control v.8 Fuzzy Sets L. A. Zadeh
  2. IEEE Trans. on Fuzzy Systems v.3 no.1 A fuzzy-logic based approach to qualitative modeling M. Sugeno;T. Yasukawa
  3. Neuro-Fuzzy and Soft Computing: A Computational Approach to Learning and Machine Intelligence J. S. R. Jang;C. T. Sun;E. Mizutani
  4. Neural Fuzzy Systems: A Neuro-Fuzzy Synergism to Intelligent Systems C. T. Lin;C. S. G. Lee
  5. A Course in Fuzzy Systems and Control L. X. Wang
  6. IEEE Trans. on Fuzzy Systems v.6 no.4 A Transformed Input-Domain Approach to Fuzzy Modeling E.T. Kim;M.K. Park;S.W. Kim;M.G. Park
  7. IEEE Trans. on System, Man, and Cybern. v.23 no.3 ANFIS : Adaptive-Networks-based Fuzzy Inference System J. S. R. Jang
  8. IEEE Trans. on Fuzzy System v.2 no.1 Reinforcement structure parameter learning forneural network-based fuzzy logic control system C. T. Lin;C. S. G. Lee
  9. Independent Component Analysis : Theory and Applications T. W. Lee
  10. 한국퍼지 및 지능 시스템학회 논문지 v.9 no.3 정수장 응집제 주입공정의 적응네트워크기반 퍼지 시스템 모델링 곽근창;한태환;유정웅;전명근
  11. Time Series Analysis, Forecasting and Control G. E. P. Box;G. M. Jenkins
  12. Fuzzy Sets and Systems v.4 The evaluation of fuzzy models derived from experimental data R.M. Tong
  13. IEEE Trans. on Systems, Man, and Cybern. v.17 Fuzzy model identification and self-learning for dynamics systems C. W. Xu;Y. Z. Lu
  14. Fuzzy Sets and Systems v.13 An identification algorithm in fuzzy relational systems W. Pedrycz
  15. Fuzzy Sets and Systems v.86 Fuzzy system modeling by fuzzy partition and GA hybrid schemes Y. H. Joo;H. S. Hwang;K. B. Kim;K. B. Woo
  16. IEEE Trans. on Fuzzy Systems v.3 Building sugeno-type models using fuzzy discretization and orthogonal parameter estimation techniques L. Wang;R. Langari
  17. Fuzzy Sets and Systems v.42 Successive identification of a fuzzy model andits applications to prediction of a complex system M. Sugeno;K. Tanaka
  18. IEEE Trans. on Fuzzy Systems v.3 A new approach to fuzzy-neural modeling Y. Lin;G. A. Cunningham
  19. IEEE Trans. on Fuzzy Systems v.5 A new approach to fuzzy modeling E. Kim;M. Park;S. Ji;M. Park
  20. IEEE Trans. on Fuzzy Systems v.6 A Transformed Input-Output Approach to fuzzy modeling E. Kim;M. Park;S. Ji;M. Park
  21. Neural Networks v.10 Stability analysis of learning algorithms for blind source separation S. Amari;T.P. Chen;A. Cichocki
  22. In Advances in Neural Information Processing v.8 A new learning algorithm for blind source separation S. Amari;A. Cichocki;H.H. Yang