結合區間第二型模糊非對稱歸屬函數及遞迴類神經網路系統之研究與應用

The Study and Applications of Combining Interval Type-2 Fuzzy Asymmetric Membership Functions and Recurrent Neural Network System

胡子偉(Tzu-Wei Hu)

指導教授 ： 李慶鴻

元智大學/電機通訊學院/電機工程學系/碩士(2008年)

https://doi.org/10.6838/YZU.2008.00286

摘要

本論文提出一個具有非對稱歸屬函數之第二型模糊遞迴類神經網路(recurrent interval type-2 fuzzy neural network with asymmetric membership functions, RT2FNN-A)，網路架構主要有五層，分別是由四層前饋式網路(feedforward network)與遞迴層連結至歸屬函數層所組成，每一個第二型模糊非對稱歸屬函數，是由四個高斯函數所建構而成，並透過李亞普諾夫穩定定理與梯度坡降原理推導參數更新法則。配合非對稱參數調整方式，可增加參數調整效率與網路近似能力。RT2FNN-A系統可藉由其遞迴結構來獲得系統動態特性並加強網路的學習性能。本文將RT2FNN-A應用於非線性系統鑑別、控制及Mackey-Glass時間序列預測與非線性通道等化器。經由模擬結果可知，本文所提出之RT2FNN-A系統，由於具有遞迴架構，與一般前饋式網路在應用上，能夠使用較簡單之網路架構與較少之規則數與調整參數來獲得相同的性能指標，證實了第二型模糊遞迴類神經網路有不錯的效果。

關鍵字

第二型模糊類邏輯系統 ； 非對稱歸屬函數 ； 系統辨識 ； 非線性控制 ； 通道等化

並列摘要

In this thesis, we propose a recurrent interval type-2 fuzzy neural network with asymmetric membership functions (RT2FNN-A). The RT2FNN-A uses the interval asymmetric type-2 fuzzy sets and it implements the fuzzy logic system (FLS) in a five-layer neural network structure which contains four layer forward network and a feedback layer. Each type-2 asymmetric fuzzy member function (AFMF) is constructed by parts of four Gaussian functions. The RT2FNN-A is modified from the type-2 fuzzy neural network (T2FNN) to provide memory elements for capturing the system’s dynamic information and has the properties of high approximation accuracy and small network structure (fewer rules and tuning parameters) from the simulation results. Based on the Lyapunov theorem and gradient descent method, the convergence of RT2FNN-A is guaranteed and the corresponding learning algorithm is derived. In addition, the RT2FNN-A is applied in the identification and control of nonlinear dynamic systems. Moreover, the Mackey-Glass chaotic time-series prediction and nonlinear channel equalization are also introduced to show the performance and effectiveness of RT2FNN-A system.

並列關鍵字

type-2 fuzzy logic system ； recurrent neural network ； asymmetric membership functions ； identification ； nonlinear control ； channel equalization

參考文獻

[39] P. Z. Lin and T. T. Lee, “Robust Self-organizing Fuzzy-neural Control Using Asymmetric Gaussian Membership Functions,” International Journal of Fuzzy Systems, Vol. 9, No. 2, pp. 77-86, 2007.

[1] J. F. Baldwin and S. B. Karake, “Asymmetric Triangular Fuzzy Sets for Classification Models,” Lecture Notes in Artificial Intelligentce, Vol. 2773, pp. 364-370, 2003.

[3] O. Castillo and P. Melin, “Adaptive Noise Cancellation Using Type-2 Fuzzy Logic and Neural Networks,” IEEE International Conf. on Fuzzy Systems, Vol. 2, pp. 1093-1098, 2004.

[4] G. Chen and X. Dong, “Identification and Control of Chaotic Systems: An Artificial Neural Network Approach,” IEEE International Sym. on Circuits and Systems, Vol. 2, pp. 1177-1182, 1995.

[6] S. Chen, B. Mulgrew, and S. McLaughlin, “A Clustering Technique for Digital Communication Channel Equalization Using Radial Basis Function Network,” IEEE Trans. on Neural Networks, Vol. 4, No. 2, pp. 570-579, 1993.