標題: 以頻域設計伺服馬達之共振抑制與精密控制實現
Realization of Resonance Suppression and Precision Control via the Frequency-Domain Design for Servo Motion Control Systems
作者: 徐煒智
Hsu, Wei-Chih
徐保羅
Hsu, Pau-Lo
電控工程研究所
關鍵字: 馬達;震動;共振;系統鑑別;motor;vibration;resonance;system identification
公開日期: 2008
摘要: 本論文發展以高精度交流伺服馬達為驅動端,具撓性機械結構為負載端之伺服定位控制,以伺服馬達編碼器位置為回授資訊,發展機械系統鑑別、共振與振動抑制及監控技術,並以撓性臂為實驗驗證平台,配合撓性臂頂端裝置的加速度計與攝影機擷取影像,驗證發展系統鑑別、共振與振動抑制及監控機能,在實驗上並利用本論文研製設計交流伺服馬達驅動器電力級,搭配本實驗室開發的數位控制板,達到與使用商用驅動器電力級相同抑制效能,驗證所發展方法於不同平台之可用性,以及研製設計的交流伺服馬達驅動器電力級實用性。 基於共振與振動資訊鑑別之目的,發展改良複合頻率序列(compound frequency sequence)系統鑑別方法,結合頻譜掃描與ARMAX參數估測系統鑑別優點,達成系統共振與振動頻率與阻尼比鑑別之目的。 在伺服馬達速度與位置控制中,利用複合頻率系統鑑別方法所得資訊,分別設計一帶斥濾波器(notch filter)與前置帶斥濾波器(feed-forward notch filter)以抑制於速度迴路中高頻共振以及位置迴路撓性臂低頻振動,提升於速度迴路系統運作頻寬以及縮短位置控制安定時間。
This Thesis develops suppression techniques for both the high-frequency resonance and the low-frequency vibration on a flexible arm implemented on a permanent magnet synchronous servo motor. The system identification, vibration suppression, and monitoring are developed in this Thesis based on the output of the encoder and the current. A power stage for the motor driver combined with the DSP-based digital control board has been developed in this Thesis. The compound frequency sequence identification method concerning the benefits of sine-sweep frequency approach and the parameter estimation time approach is developed for identification of resonance in the velocity loop and the vibration in the position loop, separately. The high-frequency resonance and the low-frequency vibration are suppressed by the design of notch filter and the feed-forward notch filter, separately. With the proposed method, the servo motor with the flexible arm can be operated with a higher bandwidth in velocity control and the settling time in position is also significantly reduced. Images captured from the camera mounted on the end of the flexible arm have proven the successful vibration suppression of the proposed method.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079612556
http://hdl.handle.net/11536/41872
顯示於類別:畢業論文


文件中的檔案:

  1. 255601.pdf

若為 zip 檔案,請下載檔案解壓縮後,用瀏覽器開啟資料夾中的 index.html 瀏覽全文。