具有非對稱高斯函數的適應式自我建構模糊類神經網路之研究

STUDY ON ADAPTIVE SELF-CONSTRUCTING FUZZY NEURAL NETWORK USING ASYMMETRIC GAUSSIANMEMBERSHIP FUNCTIONS

江明峰(Ming-Feng Chiang)

指導教授 ： 呂虹慶

大同大學/電機資訊學院/電機工程學系所/碩士(2008年)

摘要

本文提出了具有非對稱高斯函數的自我建構模糊類神經網路控制器。非對稱高斯函數被利用去提升自我建構非對稱高斯函數模糊類神經網路的學習能力及靈敏性。所提出適應自我建構非對稱模糊類神經網路滑動模式控制系統是有計算控制器和強健控制器所組成。適應自我建構非對稱模糊類神經網路被利用為一個主要的控制器, 自我建構非對稱模糊類神經網路辯證器被設計去及時估測系統的動態。強健控制器不僅被利用去衰減實際的非線性系統與近似自我建構非對稱模糊類神經網路模組之間的近似誤差,而且被建立去估測系統不確定性的邊界。馬氏距離在本篇論文被當作為是否產生或消除神經元的準則。及時適應法則是基於里亞普諾夫推導而來,因此能夠保證系統的穩定度。最後,模擬結果證實了所提出的適應自我建構非對稱高斯函數模糊類神經網路控制器之性能及有效性。

關鍵字

自我建構模糊類神經網路 ； 強健控制 ； 參數估測 ； 滑動模式控制

並列摘要

The self-constructing fuzzy neural network using asymmetric Gaussian membership function (SCAFNN) controller is proposed in this thesis. The asymmetric Gaussian membership function is utilized to upgrade learning capability and flexibility of SCAFNN. The proposed adaptive SCAFNN sliding-mode control system comprises a computation controller and a robust controller. The adaptive SCAFNN system is utilized as the principal controller, in which an SCAFNN estimator is designed to estimate the parameter of system dynamics on-line. The robust controller is not only utilized to attenuate the effects of approximation error between the real nonlinear system and an approximate SCAFNN dynamics but also developed to estimate the uncertainty bound. Mahalanobis distance (M-distance) method in this thesis is employed as the criterion to identify the neurons will be generated / eliminated or not. The on-line adaptive laws are derived based on the sense of Lyapunov so that stability of the system can be guaranteed. Finally, the simulation results of the examples are provided to demonstrate the performance and effectiveness of the proposed controller.

並列關鍵字

self-constructing asymmetric fuzzy neural networ ； robust control ； system parameter estimation ； sliding-mode control

參考文獻

[1] J. M. Mendel, A Prelude to Neural Networks: Adaptive and Learning Systems. Englewood Cliffs, NJ: Prentice-Hall, 1994.

[3] K. S. Narendra and K. Parthasarathy, “Identification and control of dynamical systems using neural networks,” IEEE Transaction on Neural Networks, vol. 1, no. 1, pp. 4-27, Mar. 1990.

[4] C. M. Lin and C. F. Hsu, “Neural-network-based adaptive control for induction servomotor drive system,” IEEE Transaction on Industrial Electronics, vol. 49, no. 1, pp. 115-123, Feb. 2002.

[5] C. M. Lin and C. F. Hsu, “Neural-network hybrid control for antilock braking systems,” IEEE Transactions on Neural Networks, vol. 14, no. 2, pp. 351-359, Mar. 2003.

[6] C. C. Ku and K. Y. Lee, “Diagonal recurrent neural networks for dynamic systems control,” IEEE Transactions on Neural Networks, vol. 6, no. 1, pp. 144-156, Jan. 1995.