本研究利用直接轉矩控制(direct torque control, DTC)理論來實現感應馬達控制,其相較於磁場導向控制的優點為動態響應快速、計算量較少且系統架構簡單。 本研究結合小腦模型控制器與模糊控制理論設計出模糊小腦模型控制器(FCMAC),並依據Lyapunov定理推導出適應性模糊小腦控制器權重更新法則,以確保系統之穩定性。本研究也根據適應性定子磁通估測器(ASFO)架構來設計轉速估測器,並將模糊控制器植入速度估測器中,以實現不需破壞感應馬達結構的無速度感測器控制。另外,由於感應馬達在長時間的運轉下,其定子電阻值會因為溫升造成變動,因此本研究在定子電阻估測上也根據適應性定子磁通估測器架構推導出適應性定子電阻估測器來調適定子電阻變化。 經模擬及實驗結果證明,將適應性模糊小腦模型轉速控制器與適應性定子磁通估測器植入感應馬達直接轉矩控制系統中,在馬達負載轉矩為 ,轉速控制範圍在36 rpm至1800 rpm時,皆具有優異的轉速動態響應以及能夠準確的估測出轉速和定子電阻。所設計之適應性定子電阻估測器在定子電阻變動時也可準確的估測出電阻值。
The direct torque control (DTC) is implemented in induction motor drives in the thesis. Compared with the FOC, the advantages of the DTC include fast dynamical responses, low computation complexity and simple structure. This thesis designs the fuzzy cerebellar model articulation controller (FCMAC) based on cerebella model articulation controller (CMAC) and fuzzy control theory, and derives the adaptation law by Lyapunov theory to ensure the stability of the system. This scheme designs speed estimator from an adaptive stator flux observer (ASFO) and using fuzzy logic controller (FLC) into it to achieve speed sensorless control. In addition, it is well known that the stator resistance of induction motor varies with temperature rising. This thesis also designs stator resistance estimator from an adaptive stator flux observer to adapt to stator resistance variation. As demonstrated in simulated and experimental results, it is observed that excellent speed tracking performance and the accurate estimation of speed and stator resistance are shown in wide speed range (36 rpm-1800 rpm). In addition, the adaptive stator resistance estimator in this thesis can estimate stator resistance accurately.