本文結合遞迴小腦模型控制器及適應性理論並以小波函數為基底函數,設計出適應性遞迴小波小腦模型控制器(adaptive recurrent wavelet cerebellar model articulation controller, ARWCMAC)。本文提出的ARWCMAC比傳統的小腦模型控制器有更好的學習率以及動態響應。此外,權重更新式學習可使追蹤的速度誤差快速收斂。使用Lyapunov來決定ARWCMAC的學習規則,使此控制器在系統上是穩定的。 本文提出的速度控制器整合了模糊轉矩和磁通控制器、模糊定子電阻估測器於感應馬達直接轉矩控制(direct torque control, DTC),且本文所提出的速度控制器整合了Takagi-Sugeno (T-S) 模糊磁通估測器與適應性Takagi-Sugeno-Kang (TSK) 模糊定轉子電阻估測器、TSK模糊轉速估測器於感應馬達磁場導向控制(field oriented control, FOC)。 最後,本文提出的速度控制器分別結合了直接轉控制架構以及磁場導向控制架構。經模擬結果證明,在馬達負載轉矩為8 Nm,轉速控制範圍在36 rpm至2000 rpm時,所提出方法皆具有優異的轉速動態響應。
In this study, the adaptive recurrent wavelet cerebellar model articulation controller (ARWCMAC) is proposed, which is designed based on the adaptive control theory and the Recurrent Cerebellar Model Articulation controller (CMAC) with wavelet basis function. The proposed ARWCMAC has superior capability to the conventional cerebellar model articulation controller in efficient learning mechanism and dynamic response. Furthermore, the variable optimal learning-rates are derived to achieve the fastest convergence of tracking speed error. The analytical method based on a Lyapunov function is proposed to determine the learning rates of ARWCMAC, so that the stability of the system can be guaranteed. This proposed speed controller integrated FTC, FFC and FSRE for DTC of induction motor, and use an ARWCMAC integrated T-S fuzzy flux estimator, TSK rotor resistance estimator and TSK estimator for FOC of induction motor. Finally, this propose speed controller integrated DTC, FOC structure for induction motor.Acording to the simuliaton, the proposed direct torque control systems have excellent speed response and robustness within 36 rpm to 2000 rpm during 8 Nm load torque.