摘要
本論文採用強健控制，將量化回授理論應用於ALMA計畫中的nutator系統。Nutator為多輸入多輸出的系統，系統目的在於消除天文反射望遠鏡觀察太空物體時所接收到的背景雜訊。運作的環境位於5000公尺高海拔處的Atacama沙漠，溫差超過攝氏40度，且環境陣風速度高達每秒20公尺，工作環境十分嚴苛。nutator的規格要求極高且追求精確，穩定時間需小於0.01秒、上升時間小於0.005秒、最大上升超越量小於10%。雖然傳統的PID控制器也可以達到所要求的表現規格，但在系統因環境變數而產生參數變化時，便無法維持系統的強健性。相反的，在經過模擬與實驗，量化回授理論不但能夠達到規格要求，更表現出傳統PID控制器所不具備的強健特性，明顯提升nutator於真實運作環境中所必須擁有的強健控制性能。
並列摘要
This thesis presents a robust control method, Quantitative Feedback Theory (QFT) to the Atacama Large Millimeter Array (ALMA) nutator. Nutator, a Multi-Input Multi-Output (MIMO) system and is used on an astronomical telescope to remove the background noise while observing objects in the space. The nutator works in a harsh environment: large temperature variation within -20℃ to 20℃, the 5000 meters high altitude and the wind speed of 20 meters per second. The requirements of the specification are expected to maintain in a high level: the 0.01 second of the settling time, 0.005 second of the rise time and the maximum overshoot should be less than 10%. Though the traditional PID control may achieve the specification requirements, but it may not be able to maintain the system performance when the parameter variation happens. However, the QFT control method does reach a stronger robustness than the PID control method. The QFT control can not only achieve the performance requirements but also fulfill the robustness characteristic that the nutator needs in the real working environment.