自動化製程管制系統之建置

Implementation of Automatic Process for Shop Floor Control Systems

胡正修(Cheng-Hsiu Hu)

指導教授 ： 陳文輝

國立臺北科技大學/機電學院/自動化科技研究所/碩士(2013年)

https://doi.org/10.6841/NTUT.2013.00488

摘要

自動化製程管制系統是工廠運作之最佳利器，亦是決定製造產業獲利與提升競爭力的重要因素之一。自動化製程管制系統要設定目標除監控生產線之現場製程能力，確保產出品質與效率外，亦可履行自動化製程資訊擷取與整合，培養如何應用資料庫分析製程問題，進而獲得有效解決問題之方法。然而，在工業電腦產業乃隸屬多樣少量之生產型態，於製程中之資訊有限，且需在極短時間內完成交貨，如何迅速、正確地設定製程參數，找出製程問題點，並提出有效改善對策，值得深入探討。如將工業電腦之生產製造，以傳統之統計製程管制方法來控制製程能力，實難達到預期的改善目標。因此，本研究方法試建置自動化製程管制系統，蒐集各製程資訊儲存為資料庫，這些資訊包括各專家個案分析結果等，並開發關聯式資料庫程式設計，歸納各製程要素成相對應因果關係，執行系統化電腦管理模式，另使用貝氏網路之機率圖形，將特定領域中之不確定性組合成聯合機率，幫助有效找出問題點與改善對策。
本研究進行分別以板卡組裝廠之生產線一、二為例，並以製程直通率為依據做為產線對製程能力的執行成效與問題解決能力來做分析，實驗結果顯示生產線一使用了自動化製程管制系統對於提升製程能力及問題解決能力明顯優於生產線二之傳統式統計製程管制方法。

關鍵字

製造執行系統 ； 統計製程管制 ； 貝氏網路 ； 失效模式與效應分析

並列摘要

Automatic process control system is the best instrument for operation of a plant and an important determinant for profit of manufacture industry and promotion of competitiveness. It not only sets a goal to monitor process ability of the production line and ensure product quality and efficiency, but also implements extraction and integration of automatic production data, fosters the ability of using database to solve problems, and figures out methods to solve the problems effectively. However, the industry of Industrial Personal Computer (IPC) is contributed to a production type of high variety but low volume. It generates limited data during the process but requires extremely short delivery lead time. It is worth further exploring how to set process parameters rapidly and correctly, figures out the problems of the process, and proposes effective corrective countermeasure. Controlling the process ability of production of IPC with traditional statistical process control would hardly accomplish the participated goal. Hence, the proposed approach attempts to establish an automatic process control system to collect data from each process including such as result of cases analyzed by each specialist and so establish a database, develop relational database program design, generalize elements of each process into corresponding causality, and execute systemized computer management mode. It further uses the probability and graphic pattern of Baysian network to implement uncertainties of specific areas into joint probability, which is beneficial for identifying problems and countermeasures.
This study takes Production Line I and II of a board assembly factory as a case and analyzes execution efficacy and problem-solution ability in process capacity of the production lines based on process First Pass Yield (FPY). Experiment results show that Production Line I adopting automatic process control system has significantly better process promotion ability and problem-solving ability than Production Line II adopting traditional statistical process control.

並列關鍵字

Manufacturing Execution System ； Statistical Process Control ； Bayesian Network ； Failure Mode and Effects Analysis

參考文獻

[1] 黃芃茨，不完美生產系統下考量產品不良率與設備可靠度之雙構面維護模式，碩士論文，國立成功大學工業與資訊管理學系研究所，台南，2011。

[3] 蔡尚穎，結合實驗設計與統計製程管制改善生產製程-以A公司為例，碩士論文，國立屏東科技大學企業管理研究所，屏東，2009。

[9] 黃崇誌，應用失效模式與效果分析及灰色關聯分析於製程改善，碩士論文，國立成功大學工學院工程管理專班，台南，2008。

[13] 林子鐸，一個以貝氏網路為基礎的風險評估工具，碩士論文，國立台北科技大學資訊工程研究所，台北，2007。

[20] 蔡聰男，自適應式表面黏著製程品質預測控制系統之發展，博士論文，國立成功大學製造工程研究所，台南，2003。