| 摘要: | 由於半導體微控制器的大幅精進，所以使得原本無法輕易取得的大量即時運算，變得輕而易舉。因而造成需要大量計算的磁場導向控制，日益普及。因為此種驅動與控制方式可將馬達性能充分發揮，所以磁場導向控制在今日的馬達驅動上，成為必備條件。本論文針對交流馬達的二大分支(感應與同步馬達)進行理論分析，並陳述共同與相異的部分。繼而，在解析各自傳統磁場導向控制的過程中，發現理論上的盲點。進一步，提出修正方法以便得到更好的馬達驅動性能。在此二分支的馬達上進行磁場導向控制的首要工作，均是定位出轉子磁通的位置。
對於交流感應馬達而言，轉子磁通是在運轉過程中，由定子磁場感應而生。所以不是一個固定存在的物理量。若利用外加感測器，則會因為技術與成本的瓶頸，破壞此類馬達的普及性。而傳統的驅動方式，又有許多理論上的盲點；諸如間接磁場導向(IFOC)計算轉子磁通角度是根據定子電流命令，而非回授，造成驅動失真且效率的低落。因而，本文在此的目標是在不增加任何製造成本的前提下，提出可以消除理論盲點的方法。所以本文建立一套自適應的機制,能在穩態運轉時，直接以演算取出準確的轉子磁通位置，並以鎖相迴路(PLL)維持在穩態的情況下，將此位置逐步代入系統中。一旦穩態破壞時，能夠退出，並取回傳統IFOC，以待再次進入穩態後，套用筆者提出的新方法，以得到更低的操作電流。
此外，對於交流同步馬達而言，雖然具有實際存在的固定轉子磁通，但是在組裝的過程中，會因為端蓋組裝而與外部的角度編碼器失準，此外，編碼器是一個昂貴且耗損率極高的裝置。再者，馬達的安裝位置通常位於設備的核心處，若需要定期的更換，勢必大幅地造成工業界的困擾。所以無位置感測器的磁場導向控制若能達成準確可靠的條件，實為工業界所期盼。不過本文發現，傳統的無感測器磁場導向控制，在理論架構上，亦有尚未被發掘的缺陷；諸如估算結果端賴預知的系統參數，並且無法隨著操作條件自行調整，所以在此致力找出可行且不增加成本的解決方法。;With the significant advancement of semiconductor microcontrollers, real-time computational capabilities that were once difficult to achieve have become increasingly accessible. As a result, field-oriented control (FOC), which demands extensive computation, has become widely adopted due to its ability to fully exploit motor performance. Today, FOC is considered essential in modern motor drive systems.
This paper presents a unified theoretical analysis of the two main branches of AC motors—induction motors (IMs) and synchronous motors (SMs)—highlighting both their commonalities and distinctions. Through a detailed examination of the traditional FOC schemes for each motor type, several theoretical limitations are identified. To address these, improved control strategies are proposed to enhance drive performance without increasing hardware cost. Central to both motor types is the precise estimation of rotor flux position, which is critical for effective FOC implementation.
For IMs, the rotor flux is not a static physical quantity but is instead induced by the stator field during operation. The use of external sensors is often limited by cost and complexity, potentially undermining the advantages of IM-based systems. Conventional approaches, such as indirect field-oriented control (IFOC), estimate the rotor flux angle based on current command values rather than actual feedback, leading to distortions and reduced efficiency. To overcome this, a novel adaptive mechanism is developed, capable of directly estimating the rotor flux angle during steady-state operation. This mechanism integrates a phase-locked loop (PLL) to continuously track and inject the estimated angle into the system. When steady-state conditions are lost, the method automatically reverts to traditional IFOC and resumes the proposed algorithm once steady-state is re-established, yielding improved efficiency and reduced current consumption.
For SMs, although the rotor flux is inherently fixed, mechanical misalignment during assembly often leads to errors between the rotor flux position and the external encoder. Furthermore, encoders are costly and prone to wear, and their maintenance poses practical challenges in industrial environments. Therefore, reliable sensorless FOC is highly desirable. However, this study identifies key flaws in existing sensorless FOC methods, particularly their reliance on predetermined motor parameters and lack of adaptability to changing operating conditions. Accordingly, this paper proposes a low-cost, parameter-insensitive alternative that enhances sensorless performance in SM drives. |
