應用 DBSCAN 演算法以提升生產線 UPH 驗證準確率

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：97

、訪客IP：3.133.139.164

姓名

徐立宇(LI YU HSU) 查詢紙本館藏

畢業系所

工業管理研究所在職專班

論文名稱

應用 DBSCAN 演算法以提升生產線 UPH 驗證準確率

相關論文

★ 以類神經網路探討晶圓測試良率預測與重測指標值之建立	★ 六標準突破性策略—企業管理議題
★ 限制驅導式在製罐產業生產管理之應用研究	★ 應用倒傳遞類神經網路於TFT-LCD G4.5代Cell廠不良問題與解決方法之研究
★ 限制驅導式生產排程在PCBA製程的運用	★ 平衡計分卡規劃與設計之研究-以海軍後勤支援指揮部修護工廠為例
★ 木製框式車身銷售數量之組合預測研究	★ 導入符合綠色產品RoHS之供應商管理-以光通訊產業L公司為例
★ 不同產品及供應商屬性對採購要求之相關性探討－以平面式觸控面板產業為例	★ 中長期產銷規劃之個案探討 -以抽絲產業為例
★ 消耗性部品存貨管理改善研究-以某邏輯測試公司之Socket Pin為例	★ 封裝廠之機台當機修復順序即時判別機制探討
★ 客戶危害限用物質規範研究-以TFT-LCD產業個案公司為例	★ PCB壓合代工業導入ISO/TS16949品質管理系統之研究-以K公司為例
★ 報價流程與價格議價之研究–以機殼產業為例	★ 產品量產前工程變更的分類機制與其可控制性探討-以某一手機產品家族為例

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 (2026-7-31以後開放)

摘要(中)

本研究探討了應用 DBSCAN 演算法來提升生產線 UPH（每小時產量）驗證準確
率的有效性。電路板（PCB）作為電子產品中的核心部件，其需求急劇增長，對於製
造公司來說，準確預測和監控生產線產能至關重要。傳統的數據分析方法，例如 Tukey
方法，雖然在處理離群值方面有一定成效，但是在面對數據有多峰分佈和非對稱數據
時，往往會導致計算結果不夠精確。因此，本研究基於多項數據演算法的評估下，提
出了一種基於 DBSCAN（密度聚類演算法）的數據分析方法，目的在於改進現有的數
據驗證方法，提高生產線產能的計算準確性。
在研究過程中，我們首先介紹了 PCB 生產線的運作模式及相關產能計算的基本概
念，接著回顧了現有的數據分析方法及其局限性，並詳細比較了多種分群演算法，最
終選擇 DBSCAN 演算法作為研究主軸。通過對實際生產數據的測試和評估，新方法能
夠更準確地識別和排除離群值，從而提高數據驗證的準確性和可靠性。研究結果顯
示，與傳統方法相比，應用 DBSCAN 演算法能顯著提高 UPH 的驗證準確性，對於製
造公司的生產計劃和決策具有重要的指標意義。

摘要(英)

This study explores the effectiveness of applying the DBSCAN algorithm to improve the
accuracy of UPH (units per hour) verification in production lines. As printed circuit boards
(PCBs) serve as the core components in electronic products, their demand has surged
dramatically. For manufacturing companies, accurately predicting and monitoring production
line capacity is crucial. Traditional data analysis methods, such as the Tukey method, while
somewhat effective in handling outliers, often lead to imprecise results when dealing with
multi-modal and asymmetrical data distributions. Therefore, this study proposes a data
analysis method based on the DBSCAN (Density-Based Spatial Clustering of Applications
with Noise) algorithm, evaluated against multiple data algorithms, to improve existing data
verification methods and enhance the accuracy of production line capacity calculations.
During the research process, we first introduced the operational model of PCB
production lines and the basic concepts of related capacity calculations. Then, we reviewed
the existing data analysis methods and their limitations, and thoroughly compared various
clustering algorithms, ultimately selecting the DBSCAN algorithm as the main focus of the
study. Through testing and evaluating actual production data, the new method can more
accurately identify and exclude outliers, thereby improving the accuracy and reliability of data
verification. The research results indicate that, compared to traditional methods, applying the
DBSCAN algorithm significantly enhances the accuracy of UPH verification, providing
crucial insights for manufacturing companies′ production planning and decision-making.

關鍵字(中)

★ 演算法
★ 聚類演算法
★ 分群
★ 四分位數

關鍵字(英)

★ DBSCAN
★ UPH
★ K-means
★ OPTICS
★ PCB
★ Tukey

論文目次

目錄
摘要.............................................................................................................................i
ABSTRACT ...............................................................................................................................ii
誌謝 ...........................................................................................................................iii
圖目錄...........................................................................................................................vi
表目錄..........................................................................................................................vii
一、緒論................................................................................................................... - 1 -
1-1 研究背景與動機............................................................................................... - 1 -
1-2 研究目的........................................................................................................... - 1 -
1-3 研究範圍與限制............................................................................................... - 2 -
1-4 研究架構........................................................................................................... - 3 -
二、文獻回顧........................................................................................................... - 5 -
2-1 PCB 製程介紹................................................................................................... - 5 -
2-2 產能計算說明................................................................................................... - 8 -
2-2-1 Tact Time 的定義及說明................................................................................. - 8 -
2-2-2 UPH .................................................................................................................. - 9 -
2-2-3 評估指標 .......................................................................................................... - 9 -
2-3 現況分析說明................................................................................................. - 10 -
2-3-1 離群值的定義 ................................................................................................ - 10 -
2-3-2 現況數據分析法 ............................................................................................ - 11 -
2-4 現況實際數據分析........................................................................................ - 13 -
三、研究方法......................................................................................................... - 16 -
3-1 各類數據分析法的比較................................................................................. - 17 -
3-2 分群演算法的比較......................................................................................... - 18 -
3-2-1 分割式演算法 ................................................................................................ - 18 -
3-2-2 階層式演算法 ................................................................................................ - 19 -
3-2-3 密度式分群演算法 ........................................................................................ - 19 -
3-2-4 網格式分群演算法 ........................................................................................ - 20 -
3-3 DBSCAN 演算法............................................................................................ - 21 -
3-3-1 DBSCAN 演算法說明 ................................................................................... - 21 -
3-3-2 DBSCAN 參數設定與步驟 ........................................................................... - 22 -
3-3-3 分群演算法的參數驗證 ................................................................................ - 24 -
3-3-4 DBSCAN 演算程式碼 ................................................................................... - 27 -
3-4-5 DBSCAN 演算程式碼結果的判讀 ............................................................... - 29 -
v
四、實際應用與分析............................................................................................. - 30 -
4-1 現況驗證結果說明........................................................................................ - 30 -
4-2 數據分析方法過程嘗試結果........................................................................ - 35 -
4-2-1 K-means 聚類演算法的測試....................................................................... - 35 -
4-2-2 OPTICS 分群演算法的測試........................................................................ - 38 -
4-3 DBSCAN 分析結果....................................................................................... - 40 -
4-4 各研究方法比較............................................................................................ - 42 -
五、結論與建議..................................................................................................... - 45 -
5-1 結論................................................................................................................ - 45 -
5-2 研究貢獻........................................................................................................ - 46 -
5-3 研究限制的突破............................................................................................ - 46 -
5-4 未來研究建議................................................................................................ - 47 -
參考文獻 ..................................................................................................................... - 48 -
附錄一 ..................................................................................................................... - 49 -
1. 編號 22 生產線分析..................................................................................... - 49 -
2. 編號 27 生產線分析..................................................................................... - 50

參考文獻

[1] TPCA 台灣電路板協會.(2022). TPCA Show 2022 Reveals How Taiwan’s PCBs are
Going Global Through Cutting-edge Technology and Sustainability.
[2] 白蓉生.(2020) 電路板實戰技術語手冊(科技圖書出版有限公司)
[3] Aggarwal, C. C., & Yu, P. S.** (2001). Outlier detection for high dimensional data.
*SIGMOD Record (ACM Special Interest Group on Management of Data)*, 30(2),
page.37-46.
[4] Hoaglin, D. C., Mosteller, F., & Tukey, J. W. (1983). Understanding Robust and
Exploratory Data Analysis.
[5] Iglewicz, B., & Hoaglin, D. C. (1993). How to Detect and Handle Outliers. *ASQC
Quality Press*.
[6] Zhang, J., Yu, X., Li, Y., Zhang, S., Xun, Y., & Qin, X. (2016). A relevant subspacebased contextual outlier mining algorithm. Knowledge-Based Systems, 99, 1-9.
doi:10.1016/j.knosys.2016.02.003
[7] 鄧翔運，2020，「基於機器學習之織布定型機與冰水主機節能分析與肇因診
斷」，碩士論文，國立中央大學，電機工程學系。
[8] 余仁傑，2023，「運用新型核密度估計演算法對時空間資料進行聚類分析」，碩
士論文，國立高雄科技大學，資訊工程學系。
[9] 蔡杰修，2017，「利用集群分析方法產製站牌班表之研究」，碩士論文，國立交
通大學，運輸與物流管理學系。
[10] dbscan - Density Based Clustering of Applications with Noise (DBSCAN) and Related
Algorithms - R package
[11] Ester, M., Kriegel, H.-P., Sander, J., & Xu, X.** (1996). A Density-Based Algorithm
for Discovering Clusters in Large Spatial Databases with Noise. *Proceedings of the
Second International Conference on Knowledge Discovery and Data Mining (KDD96)*, page.226-231.
[12] Hartigan, J. A., & Wong, M. A. (1979). Algorithm AS 136: A K-means clustering
algorithm. Journal of the Royal Statistical Society: Series C (Applied Statistics), 28(1),
page.100-108. doi:10.2307/2346830
[13] Ankerst, M., Breunig, M. M., Kriegel, H.-P., & Sander, J. (1999). OPTICS: Ordering
Points To Identify the Clustering Structure. In ACM SIGMOD Record (Vol. 28, No. 2,
page. 49-60). ACM. doi:10.1145/304181.304187

指導教授

沈國基(Gwo-Ji Sheen)

審核日期

2024-7-25

推文