以製程資料比對輔助PCB不良品診斷

Using Process Data Comparison to Assist in Diagnosis of Defective PCB’s

劉漢淮(Hann-Hwai Liou)

指導教授 ： 孫天龍博士

元智大學/工程學院/工業工程與管理學系/碩士(2000年)

摘要

印刷電路板生產過程中，常會發現一些不良品並非由單一因素所造成的，而是由兩種或兩種以上的製程因素相互影響所造成的，由多個因素所造成的不良品，不太容易判斷其要因，亦很難建立一套標準的模式可供快速的查詢及提供解決方案。當有這些不良品問題時，目前都是依賴資深員工之經驗累積來判斷及應用特性要因圖(Cause And Effect Diagram)去尋求問題之要因。如此對企業來講將會造成兩方面的困擾：(一)是時效性的困擾。(二)是經驗傳承的困擾。在企業如此競爭的社會下，以上兩方面的困擾將會造成企業體莫大的損失。為加強輔助特性要因圖之分析判斷能力，我們可以先將過去發生過的不良品之各重要製程點之製程資料建立一個資料庫。並且對目前所發生的不良品，可將此批次不良品之各重要製程點之製程資料與過去之歷史資料比對，找出相類似之製程資料，觀察所對應之歷史製程資料之生產批次是否有相同之不良品問題，若有相符，則管理人員將可參考原有之改善對策來進行此批次不良問題之改善。如此將可加強品質管理人員對特性要因圖之判斷速度與精確度，使PCB不良品要因診斷做的更確實。

關鍵字

製程資料分段 ； 製程資料比對 ； PCB不良品診斷

並列摘要

During the manufacturing processes of PCBs, some defective products are often found not because of merely one, single factor but because of two or more manufacturing factors interacting with each other. To find out the main factor of the defective products caused by many factors is not an easy task. Nor is it easy to set up a standard mold to provide a quick check-up and a solution at the same time. When these defective questions come up, they are presently solved either through the accumulated experiences of the senior staff or through the application of Cause and Effect Diagram (CAED). Hence, the enterprise will encounter two major obstacles: one is the limited time; the other is the handing down of experiences. In an aggressively competitive society, these two obstacles will cause a great lost to the enterprise. To reinforce the analytical abilities of the CAED, we may set up an information database of the many important manufacturing phases of the past defective products. By comparing every important present manufacturing phase with the previous defective ones, we can find out if there is any similarity in the manufacturing process of the present defective products. Through observing the corresponding manufacturing processes, should there be any similarities, the managerial staff can take the original improving policy into consideration in making the present defective manufacturing processes better. In doing so, the managerial staff can make a quick judgment of the CAED and execute it precisely to make the analysis

並列關鍵字

Process data segment ； Process data comparison ； defective PCB’s diagnosis

參考文獻

郭玟琳，”以資料視覺輔助產能規劃的決策”，元智大學工業工程研究所碩士論文，民國88年。

Agazzi Oscar E. and Shyh-Shiaw Kuo, 1993, ”Hidden Markov Model Based Optical Character Recognition in the Presence of Deterministic Transformations”, patten recognition, Vol.26, No.12, pp.1813-1826

Jayaram K. Udupa and Ivaturi S. N. Murthy, 1980,”Syntactic Approach to ECG Rhythm Analysis”, IEEE Transactions on Biomedical Engineering, Vol.Bme-27, No.7, july, pp.370-375

Lawrence R. Rabiner, 1989,”A Tutorial on Hidden Markov Model and Selected Applications in Speech Recognition”, Proceedings of the IEEE, Vol.77, No.2, February, pp.257-286

Lin Hsin-Chih, Ling-Ling Wang and Shi-Nine Yang, 1997, ”Color Image Retrieval Based on Hidden Markov Model”, IEEE Transacttions on Image Processing, Vol.6, No.2, February, pp.332-339