- 學位論文
利用柔性演算法於多重輸入多重輸出之製程管制系統
Construct MIMO Process Control System by Using Soft Computing Methods
摘要
在全球經濟環境的快速變遷的情況下，台灣產業如何因應競爭者的挑戰與增加本身的競爭力，而強化產品品質，穩定製程管制能力已是關鍵要素之一。 過去為了提升產品品質的穩定通常是採用統計製程管制技術 (Statistical Process Control；SPC) ，以製程偵測的方式來移除可能造成製程變異的可歸屬原因；進而演變使用工程製程管制 (Engineering Process Control；EPC) 來對製程的輸出進行控制，並且持續的回饋與調整其操縱變數來補償不可控制的因子對製程所造成的干擾，以維持產品輸出的品質穩定。 本研究主要目標在於針對系統中多重輸入、多重輸出 (Multiple-Input Multiple-Output；MIMO) 的製程型態，以整合統計製程管制與工程製程管制的概念為基礎，將柔性演算法 (Soft Computing；SC) 技術與統計分析技術相結合，藉由柔性演算法來模式化複雜的製程輸入與輸出關係以獲得較佳的產出結果，期以提昇製程品質。 故本研究建構一以柔性演算法之技術為預測與參數控制模式之多重輸入多重輸出製程管制系統，其包含MIMO製程子系統、量測子系統、偏移偵測與分析子系統、類神經網路製程輸出預測子系統、基因演算法參數設計最佳化搜尋子系統、參數調整子系統等六部分，並詳述各子系統之內部運作及子系統間之相互關係。 最後建構一化學機械研磨之模擬製程，驗證系統於偏移干擾下本研究所建構之製程管制系統能有效控制製程輸出，達到穩定製程輸出品質特性之目的。
並列摘要
In the trend of the times, many continue process industry such as high tech industry or chemical process industry needs more precise production system to avoid defect. The scope of theme is multiple-input and multiple-output (MIMO) process control system. Due to the multiple input control factors, the output values may shift from the target. It will make a lot of lose and failures, and it is hard to take possession of the optimal input parameters of the continued manufacturing system. This research is applying several techniques, such as engineering process control (EPC), statistical process control (SPC), feedback control and soft computing. The soft computing methods has includes neural networks (NN), genetic algorithms (GA). This MIMO process control system includes five parts: MIMO manufacturing process subsystem, measurement subsystem, disturbance detecting and analyzing subsystem, process multiple-output forecasting by NN subsystem, genetic algorithm parameter search subsystem and parameter adjustment subsystem. In the MIMO process control system, we use soft computing methods to forecast the multiple-output values of the next lot. Once the forecasts are got, we estimate the multiple-input parameters of MIMO process system by using genetic algorithm. Finally, in this study, we create the simulated process data of chemical mechanical polishing (CMP) process and use the data to validate and verify this process control system. The result shows that the MIMO process control system can reduce the variation from target.