適應性模糊小腦模型控制器於直接轉矩控制系統之設計

Design of Adaptive Fuzzy Cerebellar Model Articulation Controller for Direct Torque Control System

王順源(Shun-Yuan Wang) ； 張朝陽(Chaur-Yang Chang) ； 黃欽彥(Qin-Yan Huang) ； 曾傳蘆(Chwan-Lu Tseng)

《臺北科技大學學報》 43卷2期 (2010/12) Pp. 27-40

https://doi.org/10.29768/JNTUT.201012.0003

摘要

感應馬達直接轉矩控制相較於向量控制有架構簡單及動態響應快等優點，可用於負載快速變化之系統上。本研究利用直接轉矩控制理論實現感應馬達轉速控制，並籍由轉速估測器來實現無轉速量測器控制，以達到節省成本及避免破壞感應馬達結構等目的。
所提出的適應性模糊小腦模型控制器結合小腦模型控制器與模糊系統，再以高斯函數作為歸屬函數，依據Lyapunov定理推導出適應性模糊小腦模型控制器，並將其運用於馬達轉速控制器。此控制器具有線土學習的能力。此外，由於馬達會因為溫升效應而造成定子電阻變動，因此本研究將模糊小腦模型控制器值入參考模型適應系統中，設計出模糊小腦模型定子電阻估測器，即時調適定子電阻值，以精準估測磁通量。
本研究將適應性模糊小腦模型轉速控制器與模糊小腦模型定子電阻估測器植入感應馬達直接轉矩控制系統中。經實驗結果證明，在寬廣的轉速運轉範圍下，皆具有優異的轉速動態響應，且在定子電阻變動時電阻估測器仍可準確的估測出電阻值，以維持優異的系統強健性。

關鍵字

模糊小腦模型控制器 ； 電阻估測器 ； 直接轉矩控制

並列摘要

As comparing to the vector control scheme, the DTC implemented in induction motor drives to gain the classic merits, including simple structure, fast system responses and low computation complexity. This paper proposes speed estimator for sensorless speed control to achieve the advantages of cost-effectiveness and structure robustness.
The proposed adaptive fuzzy cerebellar model articulation controller (AFCMAC) speed controller integrates cerebella model articulation controller (CMAC) and fuzzy system with Gaussian membership functions. The AFCMAC speed controller is derived using the Lyapunov stability theory and then applied to the motor speed control. The proposed controller has the on-line learning ability. Due to temperature influences, stator resistance of induction motor varies during operation. To overcome this problem, the FCMAC and the MRAS (model reference adaptive system) are combined to design a FCMA stator resistance estimator for real-time estimating accurate stator resistance and flux linkage.
The AFCMAC speed controller and the FCMAC stator resistance estimator are implemented on DTC induction motor drives for verifying the effectiveness and feasibility. As the experimental results demonstrate, it is observed that the excellent tracking performance in wide speed range (36-2000 rpm) is achieved by using the proposed scheme. In addition, the stator resistance can be estimated accurately by the proposed FCMAC stator resistance estimator for guaranteeing the system robust performance under parameter variations.

並列關鍵字

fuzzy cerebellar model articulation controller ； Resistance Estimator ； Direct Torque Control