透過您的圖書館登入
IP:18.117.111.1
  • 期刊
  • OpenAccess

Design of Digital Simplified MIMO Internal Model Control Using Genetic Algorithms

基因演算法應用於多變數內建模式控制之研究

摘要


工業製程系統的輸出響應經常出現遲緩和無最大超越量的情況,操作者藉由實施單調開迴路步階實驗輕易地將這種輸入-輸出反應以一階時間延遲轉移函數來模擬。本文提出以一階時間延遲轉移函數作為設計多變數內建模式控制器的基礎,並搭配基因演算法作為判別一階時間延遲轉移函數參數的方法,它不須搭配任何控制器參數調準法則既可直接應用於多變數內建模式控制系統,這種方法既簡單又容易被操作者認同。為評估一階時間延遲轉移函數作為設計多變數內建模式控制器的效能,本文擬以高階且長時間延遲轉移函數作為模擬之實例,求取的結果將與多變數比例-積分控制器的結果作一比較以驗證其正確性。結果證實,以基因演算法作為判別一階時間延遲轉移函數的參數,並以此作為設計多變數內建模式控制器的基礎,輸出性能超越多變數比例-積分控制器19.71%,因此多變數內建模式控制器被認定是有效的控制器。

並列摘要


Nonovershooting and sluggish characteristics are frequently encountered in the process control. This monotone open-loop step response can be approximated as the first order plus dead time model. It is thus proposed in this paper as a simplified model to design MIMO internal model control. Genetic algorithms will be deployed to automate the design of the simplified MIMO internal model control. This technique is easy to be adopted by the operator without any tuning methodology and then can be direct to implement. To assess the output performance of the simplified MIMO internal model control, the results will be illustrated by an example of a high order plant with long time delay and compared with the centralized PI control. The results show that the simplified MIMO internal model control using genetic algorithms makes 19.71% improvement over the centralized PI control. Thus the digital simplified MIMO internal model control proves to be an effective control.

延伸閱讀