Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/2592
標題: 灰色預測批次控制器於銅化學機械研磨之應用
Application of Grey Prediction Process Controller on Copper Chemical Mechanical Polishing
作者: 羅國寧
Lau, Gau-Ning
關鍵字: CMP;化學機械研磨;Grey Prediction;run-to-run process;LSE;灰色預測;批次製程;最小平方法
出版社: 機械工程學系
摘要: 
本論文結合批次製程控制、遞迴式最小平方演算法與灰色理論預測模型的技術,針對半導體製程設備無法做到線上量測的問題,提出線上之灰色批次預測製程控制系統架構。對於代表系統的真實矩陣以及製程的誤差,皆能透過線上估測系統的行為以及利用前饋的方式事先做誤差的補償,使得製程性能得以提昇,進一步達成製程的虛擬線上監控。
本文中所針對的受控系統為化學機械研磨機,由電腦模擬結果顯示,除了能將製程性能控制在百分之五的誤差範圍外,對於製程參數發生飽和時,亦能夠藉由對其它未飽和的製程參數之調整來補足控制力,使得製程性能能保以維持。
此外,本論文所提的控制架構,無須變更機台結構與增加任何其他感測器,因此有其可行性與經濟效益。除化學機械研磨製程外,本控制器亦可使用於其他批次製程上,可應用於其它產業之製程監控。

The goal of this research is to propose a grey prediction based run-to-run process control scheme that combines a grey prediction process controller and a least-squared process estimator to better control the output of the erratic and unstable dynamics of the copper chemical mechanical polishing processes. In this new process control scheme, the least-squared method is adopted to on-line identify the system's matrix. The grey prediction theorem is implemented to estimate the output error of the process in advance such that the output error can be eliminated by a feedforward control scheme.
Computer simulations and experiments demonstrate that the proposed run-to-run control structure can restrict the process error of the copper chemical mechanical polishing within 5% under variable saturation situations.
To employ the new control method, no extra sensor and machine modification is required. In addition to the CMP, others semiconductor batch processes such as the chemical vapor deposition, physical vapor deposition, doping, and etching are potential application domains.
URI: http://hdl.handle.net/11455/2592
Appears in Collections:機械工程學系所

Show full item record
 

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.