雙集結式機台包含重入流程之排程規劃

Scheduling of Two-Cluster Tools with Four Blades including Revisiting Process

張書瑋(Shu-Wei Chang)

指導教授 ： 黃漢邦

國立臺灣大學/工學院/工業工程學研究所/碩士(2013年)

https://doi.org/10.6342/NTU.2013.00064

摘要

集結式機台廣泛地應用在半導體製造上，由於半導體的製程日益複雜，多集結式機台的應用逐漸顯現出來。尤其，某些晶圓的製程，像是原子層沉積這類的重入過程，晶圓必須由相同的某些機台加工很多次，就很適合利用多集結式機台。因此，如何用有效率的方法找到一個好的排程，對生產改善與成本降低是至為重要的。
本論文針對含有4個機器手臂且包含重入過程之一個雙集結式機台進行探討，在考慮所有晶圓都依循同一加工路徑的狀況下，求出系統在穩態時的輸出與排程。由於問題的複雜度，我們使用分解法將此一問題分成兩個問題，分別是含有4個機器手臂但不包含重入過程之一個雙集結式機台問題，以及有2個機器手臂並且包含重入過程之一個單集結式機台的問題。
在探討4個機器手臂並且包含重入過程之一個雙集結式機台問題底下，我們藉由分析法來決定2個機器手臂並且包含重入過程之單集結式機台的排程。並且根據此結果，藉由將連續的重入過程之機台視為單一無重入過程之機台，建立一個含有4個機器手臂並且包含重入過程之雙集結式機台的轉換模型。然後，使用含有4個機器手臂但不包含重入過程之一個雙集結式機台問題的混和整數規劃，來求出另一2個機器手臂不包含重入過程之單集結式機台的排程。
最後，我們再用模擬系統來確認系統的有效性。從結果顯示，在1或2單位的週期循環中，可以發現含有4個機器手臂並且包含重入過程之雙集結式機台的轉換模型的結果，與含有4個機器手臂並且包含重入過程之雙集結式機台的原模型結果相同。

關鍵字

多集結式機台 ； 混和整數規劃 ； 重入過程 ； 機械手臂

並列摘要

Cluster tools are widely used as semiconductor manufacturing equipment. In particular, they are suitable for wafer fabrication processes with revisiting process, such as the atomic layer deposition (ALD) process in which the wafers need to visit some processing modules for many times. How to find an efficient way for scheduling a multi-cluster tool is critical for productivity improvement and cost reduction.
In this study, we analyze the steady-state throughput and the scheduling of the two-cluster tool with four blades including revisiting process. We consider the case where all wafers follow the same processing flow within two cluster tools. The two-cluster tool with four blades including revisiting process problem is decomposed into the two-cluster tool with four blades without revisiting process problem and the single cluster tool with two blades including revisiting process problem.
First, we apply the analyzing method to the single cluster tool with two blades including revisiting process problem. We determine the robot actions in the cluster tool including revisiting process. Based on the result, we build a transformation model of the two-cluster tool with four blades including revisiting process by regarding the consecutive revisiting processing modules as one processing module. Then, we use mixed integer programming method to find the solutions of the robot actions for the transformation model.
Finally, we use simulation system to validate its effectiveness. The result of transformation model of the two-cluster tool with four blades including revisiting process shows that it is the same as the result of the original model of two-cluster tool with four blades including revisiting process by the simulation system in the 1-unit or 2-unit cycle.

並列關鍵字

multi-cluster tools ； mixed integer programming ； revisiting process ； four blades

參考文獻

[32] Z. Y. Xiao, “Development of Remote Control System of a Semiconductor Cluster Tool,” Master Department of Mechanical Engineering, National Taiwan University, 2001.

[1] N. Brauner and G. Finke, “On a Conjecture About Robotic Cells: New Simplified Proof for the Three-Machine Case,” INFOR, Vol. 37, No. 1, pp. 20-36, 1999.

[2] W. K. Chan, J. Yi, and S. Ding, “On the Optimality of One-Unit Cycle Scheduling of Multi-Cluster Tools with Single-Blade Robots,” Proceedings of the 3rd Annual IEEE Conference on Automation Science and Engineering, Scottsdale, AZ, USA, pp. 392-397, 2007.

[4] W. K. V. Chan, S. Ding, J. Yi, and D. Song, “Optimal Scheduling of Multicluster Tools with Constant Robot Moving Times, Part Ii: Tree-Like Topology Configurations,” IEEE Transactions onAutomation Science and Engineering, Vol. 8, No. 1, pp. 17-28, 2011.

[5] W. K. V. Chan and T. M. Roeder, “On Gradient Estimation of Scheduling for Multi-Cluster Tools with General Robot Moving Times,” IEEE Conference onAutomation Science and Engineering, Marriott Eaton Centre Hotel Toronto, Ontario, Canada, pp. 112-117, 2010.