基因演算法對工具機軌跡交叉耦合預補償控制法參數之最佳化

Abstract

在本論文中，使用基因演算法作為工具，去找出數控機械雙軸向刀具輪廓控制法中的最佳的增益值，以往通常採用試誤法去處理類似問題，但當控制系統變大且複雜時，這種方式會變得困難又無效率。 數控機械刀具輪廓控制法，包括傳統循跡系統--未耦合單軸向軌跡控制法(US)，交叉耦合雙軸向控制法 (CCS)，及交叉耦合雙軸向預補償控制法(CCPM)，本論文將基因演算法應用在前述系列控制法中，成功的獲得系統最佳增益值，可證明基因演算法對動態系統增益值的規畫應用是有參考價值的。

ABSTRACT

Genetic algorithm is used as a tool to find the optimal gains for contour control of biaxial CNC machine tools in this work. In the past, this was usually done by trial and error, which became difficult and inefficient when the system became large and complicated. The addressed contour control schemes for CNC machine tools include conventional tracking system (Uncoupled system, US) , Cross-coupled biaxial control (CCS), and Cross-coupled pre-compensation control (CCPM ). The successful implementation of genetic algorithm in sequential determination of system gains for the addressed control schemes showed that the genetic algorithm is applicable to the dynamic system in a manner of system gains configuration.