透過您的圖書館登入
IP:18.191.174.168
  • 學位論文

智慧型雷射功率與加工運動平台之控制

Intelligent Laser Power and Machining Stage Motion Control

指導教授 : 陳金聖
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


本論文針對「智慧型雷射功率與加工運動平台之控制」主題進行研究。 雷射加工廣泛應用於許多工業應用,如醫療,機械,軍事。無論是雷射輸出功率和加工運動平台性能皆會影響雷射加工品質。由於雷射控制輸入與功率輸出存在高度非線性關係,不易模型化,故本論文採用智慧型自調式Fuzzy PI 控制器對雷射功率進行穩定控制。採用FRE (Fuzzy reasoning engine)之優點為可針對雷射系統產生適當的受控輸入,使複雜的雷射系統可獲得理想的響應。由實驗結果得知雷射系統搭配此智慧型自調式Fuzzy PI 控制器效果最佳。 全域型順滑控制搭配仿PID類神經網路估測器於線性馬達控制,全域型順滑模態控制器,屬於一種強健控制系統架構,保證系統動態均於順滑層內,消除一般傳統順滑控制器迫近模態,可對系統平台達到理想之控制及定位。仿PID類神經網路估測器,取代傳統的切換函數與飽和函數,透過類神經網路動態學習,對於摩擦力有一定的掌握能力,可免除切換函數之切跳現象。將運動平台的追蹤性能,比較傳統控制器的輸出結果,得知加入全域型順滑控制搭配仿PID類神經網路估測器後,可以達到最好的強健性與控制性能。

並列摘要


This paper for “Intelligent Laser Power and Machining Stage Motion Control” subjects to research. Laser machining is widely used in many industry applications, such as medical, machinery, military. Both the laser output power and performance of machining motion stage will affect the quality of laser machining. The laser system is a nonlinear system due to its complexity and easily affected by environment conditions. This paper uses intelligent self-tuning fuzzy PI controller to stabilize the laser power control. The advantages of using FRE for nonlinear controlled systems can be controlled to produce proper input, and obtain the ideal response. In addition, we adopt global sliding mode control with the PID-like neural network (GSMCNN) estimator in linear motor control. Global sliding mode controller is a robust control architecture that can guarantee the system’s responses are controlled in the sliding layer. The elimination of the traditional sliding mode controller is approaching fast switching, the system can achieve the desired control response and position. PID-like neural network estimator replaces the traditional switching function and saturation function. By neural network going to dynamic learning, we can overcome the friction can eliminate the chattering phenomenon of switching function. Finally, after comparing GSMCNN with traditional controller, we can find that GSMCNN can achieve the best robustness and control performance.

參考文獻


[1] M. Kuipers, P. Ioannou, “Multiple model adaptive control with mixing,” IEEE Transactions on Automatic Control , Cancun, 9-11 Dec. 2008, pp. 1822-1836.
[2] B.C. Kuo, Automatic Control Systems, Prentice-Hall, Inc., 1995.
[3] G. Chen, H. Ying, “Stability analysis of nonlinear fuzzy PI control systems,” Industrial Fuzzy Control and Intelligent Systems, Houston, 1-3 Dec. 1993, pp. 128-133.
[8] 張運添,「改善歸屬函數之模糊PID控制器」,碩士論文,國立中央大學電機工程研究所,2006。
[10] U. Itkis, “Control System of Variable Structure, ”New York: John Wieley, 1976.

延伸閱讀