第二型模糊類神經網路系統及其應用: 訊號處理、非線性系統鑑別及控制

Type-2 Fuzzy Neural Network System and Its Applications in Signal Processing, Nonlinear System Identification, and Control.

林育慶(Yu-Ching Lin)

指導教授 ： 李慶鴻

元智大學/工程學院/電機工程學系/碩士(2004年)

摘要

本論文提出第二型模糊類神經網路並搭配最佳的學習法則以處理系統之不確定性，此網路實現第二型模糊推論系統於多層類神經網路架構，其中第二型模糊類神經網路的前件部與後件部為第二型模糊集合，並經由型態還原與解模糊化後得到的輸出。在網路分析上，本文成功的將第一型模糊類神經網路的特性，如模糊推論引擎，廣泛的近似能力及參數收斂分析，推廣至第二型模糊類神經網路。因為型態還原的計算複雜，第二型模糊類神經網路的計算量大於第一型模糊類神經網路。關於系統存在不確定性方面，我們以第二型模糊類神經網路實現Mackey-Glass 時間序列之預測器，並且將之應用在對非線性時變通道之等化器。關於非線性控制方面，我們使用第二型模糊類神經網路建構適應性控制器。基於李亞普諾夫理論，可以嚴謹地證明出第二型模糊類神經網路的收斂能力以及其穩定。模擬結果證實了第二型模糊類神經網路的效果。

關鍵字

第二型模糊類神經網路 ； 不確定 ； 型態還原 ； 李亞普諾夫理論

並列摘要

This paper proposes a type-2 fuzzy neural network (type-2 FNN) with optimal learning algorithm to handle the system uncertainty. The type-2 FNN is inherently a multilayered connectionist network for realizing type-2 fuzzy inference using dynamic type-2 fuzzy rules. The type-2 FNN system consists of type-2 fuzzy linguistic process as the antecedent and consequent parts. The consequent part is the output through type-reduction and defuzzification. The computation of type-2 FNN is more complex than that of type-1 one due to the type-reduction. The results for the type-1 FNN- fuzzy inference engine, universal approximation, and convergence analysis are extended to the type-2 FNN. For application of signal processing, the type-2 FNN is used to be an adaptive predictor of Mackey-Glass time series, and be channel equalizer of nonlinear time-varying channel. For nonlinear control problem, we propose the adaptive control scheme using type-2 FNN systems. Based on the Lyapunov theorem, rigorous proofs are presented to guarantee the convergence of the type-2 FNN and nonlinear system stability. Several simulations are shown to illustrate the effectiveness of the type-2 FNN system.

並列關鍵字

type-2 fuzzy neural network ； uncertainty ； type-reduction ； Lyapunov theorem

參考文獻

[5] Chen, G., and X. Dong, “Identification and control of chaotic systems: an artificial neural network approach,” IEEE International Symposium on Circuits and Syst., Vol. 2, 28 April-3 May 1995.

[6] Chen, S., and B. Mulgrew, “Overcoming co-channel interference using an adaptive radial basis function equalizer,” Signal Processing, vol. 28, pp. 91-107, July 1992.

[7] Chen, S., B. Mulgrew, and S. McLaughlin, “A clustering technique for digital communication channel equalization using radial basis function network,” IEEE Trans. Neural Networks, vol. 4, no. 2, pp. 570-579, July 1993.

[8] Chen, Y. C., and C. C. Teng, “A model reference control structure using a fuzzy neural network,” Fuzzy Sets and Syst., vol. 73, pp. 291-312, 1995.

[9] Chua, L. O., M. Komuro, and T. Matsumoto, “The double scroll family: Ⅰ andⅡ,” IEEE Trans. Circuit Syst. vol. CS-33, pp. 1072-1118, 1986.