The Transactions of the Korean Institute of Electrical Engineers D (대한전기학회논문지:시스템및제어부문D)

The Korean Institute of Electrical Engineers (대한전기학회)
권호 표지

The Hybrid Multi-layer Inference Architectures and Algorithms of FPNN Based on FNN and PNN

FNN 및 PNN에 기초한 FPNN의 합성 다층 추론 구조와 알고리즘

  • 박병준 (원광대 전기전자공학부) ;
  • 오성권 (원광대 전기전자공학부) ;
  • 김현기 (수원대 전기전자정보통신공학부)
  • Published : 2000.07.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we propose Fuzzy Polynomial Neural Networks(FPNN) based on Polynomial Neural Networks(PNN) and Fuzzy Neural Networks(FNN) for model identification of complex and nonlinear systems. The proposed FPNN is generated from the mutually combined structure of both FNN and PNN. The one and the other are considered as the premise part and consequence part of FPNN structure respectively. As the consequence part of FPNN, PNN is based on Group Method of Data Handling(GMDH) method and its structure is similar to Neural Networks. But the structure of PNN is not fixed like in conventional Neural Networks and self-organizing networks that can be generated. FPNN is available effectively for multi-input variables and high-order polynomial according to the combination of FNN with PNN. Accordingly it is possible to consider the nonlinearity characteristics of process and to get better output performance with superb predictive ability. As the premise part of FPNN, FNN uses both the simplified fuzzy inference as fuzzy inference method and error back-propagation algorithm as learning rule. The parameters such as parameters of membership functions, learning rates and momentum coefficients are adjusted using genetic algorithms. And we use two kinds of FNN structure according to the division method of fuzzy space of input variables. One is basic FNN structure and uses fuzzy input space divided by each separated input variable, the other is modified FNN structure and uses fuzzy input space divided by mutually combined input variables. In order to evaluate the performance of proposed models, we use the nonlinear function and traffic route choice process. The results show that the proposed FPNN can produce the model with higher accuracy and more robustness than any other method presented previously. And also performance index related to the approximation and prediction capabilities of model is evaluated and discussed.

Keywords

References

  1. A. G. Ivahnenko, 'The group method of data handling: a rival of method of stochastic approximation,' Soviet Automatic Control, Vol. 13, No. 3, pp. 43-55, 1968
  2. T. Kondo, 'Revised GMDH algorithm estimating degree of the complete polynomial,' Tran. of the Society of Instrument and Control Engineers, Vol. 22, No. 9, pp. 928-934, 1986
  3. G. Kang and M. Sugeno, 'Fuzzy modeling,' Tran. of the Society of Instrument and Control Engineers, Vol. 23, No. 6, pp. 106-108, 1987
  4. I. Hayashi and H. Tanaka, 'The Fuzzy GMDH algorithm by possibility models and its application,' Fuzzy Sets and Systems, 36, pp. 245-258, 1990 https://doi.org/10.1016/0165-0114(90)90182-6
  5. H. Takagi and I. Hayashi, 'NN-driven Fuzzy Reasoning,' Int. J. of Approximate Reasoning, Vol. 5, NO. 3, pp. 191-212, 1991 https://doi.org/10.1016/0888-613X(91)90008-A
  6. Shin-ichi Horikawa, Takeshi Furuhashi, and Yoshiki Uchikawa, 'On Fuzzy Modeling Using Fuzzy Neural Networks with the Back Propagation Algorithm,' IEEE trans. Neural Networks, Vol. 3, No. 5, pp. 801-806, 1992 https://doi.org/10.1109/72.159069
  7. 森杉壽旁, '非集計モデルの 推定と 檢定', 日本土木計劃學講習會テキスト 15, PP. JSCE, 1984
  8. H. Nomura and Wakami, 'A Self-Tuning Method of Fuzzy Control by Descent Methods,' 4th IFSA World Conference, pp. 155-159, 1991
  9. Takwshi Yamakawa, 'A New Effective Learning Algorithm for a Neo Fuzzy Neuron Model,' 5th IFSA World Conference, pp. 1017-1020, 1993
  10. S. S. Kim, 'A Neuro-fuzzy Approach to Integration and Control of Industrial Processes : Part I,' Journal of Fuzzy Logic and Intelligent Systems, Vol. 8, No. 6, pp. 58-69, 1998
  11. Wardrop, J.G. Theoritical Aspect of Road Traffic Research, Proc. of The Institution of Civil Engineers, Part I, Vol. 1, pp. 325-378, 1952
  12. 佐左木 綱 : 道路綱における輸送計劃について, 第4回日本道路會議論文集 pp.43-46, 1957
  13. 星野 哲三 : 道路綱における交通量配分の理論, 第5回日本道路會議論文集, pp.68 sim 70, 1959
  14. Martin B.V. and Manheim, M.L., A research Program for Comparison of Traffic Assignment Techniques, Highway Research Record 88, pp. 69-84, 1965
  15. 今崎直樹, 木治潤一, 'ファジィル－ル構造化ニュ－ラネットワ－ク', 日本ファジィ學會誌 Vol.4, No.5, pp. 985-995, 1992
  16. David E. Goldberg, Genetic Algorithms in search Optimization & Machine Learning, Addison-Wesley, 1989
  17. Zbigniwe Michalewicz, Genetic Algorithms+Data Structure=Evolution Programs, Springer-Verlag, 1992
  18. S.K. Oh, and W. Pedrycz, 'Fuzzy Identification by means of Auto-Tuning Algorithm and Its Application to Nonlinear Systems,' Fuzzy Sets and Syst, Vol. 115, No. 2, pp. 205-230, 2000 https://doi.org/10.1016/S0165-0114(98)00174-2
  19. 菅野道夫 (譯:박민용, 최항식), '퍼지제어 시스템,' pp. 143-158, 大英社, 1990
  20. 오성권, 노석범, 남궁문, '퍼지-뉴럴 네트워크 구조에 의한 비선형 공정시스템의 지능형 모델링,' 한국 퍼지 및 지능시스템학회 논문지, 제5권 제4호, pp. 41-55, 1995
  21. 오성권, 윤기찬, '유전자 알고리즘과 합성 성능지수에 의한 최적 퍼지-뉴럴 네트워크 구조의 설계,' 제어자동화시스템공학논문지, 6권, 3호, pp. 273-283, 2000
  22. 박병준, '다항식 뉴럴네트워크 및 퍼지추론에 기초한 퍼지 다항식 뉴럴네트워크 구조에 관한 연구,' 원광대학교 대학원, 석사학위논문, 1999
  23. Sung Kwon Oh, Tae Chon Ahn, Witold Pedrycz, 'Fuzzy Polynomial Neural Networks-Based Structure and Its Application to Nonlinear Process Systems,' 7th IFSA World Conference, Vol. 2, pp.495-499, 1997
  24. H. S. Park, S. K. Oh, T. C. Ahn, and W. Pedrycz, 'A Study on Multi-layer Fuzzy Polynomial Inference System based on an Extended GMDH algorithm,' 1999 IEEE International Fuzzy Systems Conference Proceedings, Vol. 1, pp. 354-359, 1999 https://doi.org/10.1109/FUZZY.1999.793265
  25. 오성권, 김동원, 박병준, '다항식 뉴럴네트워크 구조의 최적 설계에 관한 연구,' 대한전기학회논문지, 49권, 3호, pp. 145-156, 2000
  26. 오성권, '퍼지모델 및 제어이론과 프로그램,' 기다리 출판사, 1999년 3월