針對非線性控制系統之自我建構第二類型模糊類神經網路設計

在本文中針對非線性控制系統提出一個自我建構第二型模糊類神經網路。所提出的自我建構第二型模糊類神經網路包含計算控制器與強建控制器兩部份。在此提出的自我建構第二型模糊類神經網路中，第二類型模糊類神經網路估測器結合第二類型模糊邏輯與類神經網路優點。此估測器被發展去估測未知的動態函數。藉著利用第二類型模糊類神經網路估測器，一個計算控制器可被獲得。此外，一個強建控制器被提出來消除不確定項。再者，藉由李普諾夫穩定理論推導出適應學習演算法。這適應學習演算法是用來線上調整第二類型模糊類神經網路估測器中的參數。一個適應誤差估測法則是用來減低強建控制器中誤差的需求。馬氏距離在本篇論文中被當作是否產生或消除神經元的準則。最後，針對非線性系統模擬結果證實了所提出的自我建構第二型模糊類神經網路控制系統的性能與有效性。

關鍵字

自我建構模糊類神經網路；第二類型模糊邏輯；滑動模式控制

並列摘要

A self-constructing type-2 fuzzy neural network (SCT2FNN) is proposed for nonlinear control system in this thesis. The SCT2FNN comprises a computation controller and a robust controller. In the proposed SCT2FNN, a T2FNN estimator, which combines the merits of a type-2 fuzzy logic system and a neural network, is developed to estimate the unknown dynamic functions. By using a T2FNN estimator, a computation controller can be obtained. Furthermore, a robust controller is proposed to confront the uncertainties. Moreover, the adaptive learning algorithms that can adjust the parameters of the T2FNN estimator on-line are derived by using the Lyapunov stability theorem. And to relax the requirement for the value of the lumped uncertainty in the robust controller, an adaptive lumped uncertainty estimation law is obtained. Mahalanobis distance method is employed as the criterion to identify that the neurons will be generated or eliminated in this thesis. Finally, the simulation results are provided to demonstrate the performance and effectiveness of the proposed control system.

並列關鍵字

sliding mode control ； self-constructing fuzzy neural network (SCFNN) ； type-2 fuzzy logic

參考文獻

generalized dynamic fuzzy neural networks,” Microprocessors and

[1]J. M. Mendel, “Fuzzy logic systems for engineering: A tutorial,”

[2]Q. Liang and J. M. Mendel, “Interval type-2 fuzzy logic systems: theory

and design,” IEEE transactions of fuzzy systems, vol. 8, no. 5, pp. 535-

550, October 2000.

國際替代計量

針對非線性控制系統之自我建構第二類型模糊類神經網路設計

未授權

主題瀏覽