| 姓名 |
陳建融(Jian-Rong Chen)
查詢紙本館藏 |
畢業系所 |
光機電工程研究所 |
| 論文名稱 |
應用神經網路智慧學習檢測技術於準直型LED曝光裝置的光源誤差探討
(Application of neural network intelligent learning and detection technology to light source error of collimated LED exposure device)
|
| 相關論文 | |
| 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 ( 永不開放)
|
| 摘要(中) |
本論文在光源光分佈的二維函數擬合曲線下藉由對稱旋轉所產生的理想三維光分佈函數曲面為基礎，透過疊合這些理想的三維光分佈函數曲面所形成的函數曲面，做為本研究用來模擬真實物理世界的三維光場的基底函數，透過調整基底函數的尺度母數與位置母數，與光學模擬軟體相比較的平方絕對誤差可以控制在 5%以下。
本論文以 UV-LED 曝光機為基礎研究對象，以三維光場分佈函數藉由準直 LED 模組單元的角度參數偏轉量的演算法，在曝光機最佳光源間距下，針對不同光源到目標平面距離 D 和光源直徑 S 的比率(距離直徑比，DSR)，DSR=10、12、14、16、18 的平面之照度均勻度皆大於 90%的情況下，透過光源排列的對稱性與隨機角度偏轉下進行照度均勻度的分辨分析，以局部光源的偏轉加上均勻度變化幅度，結合神經網路學習演算程式，從而分辨每顆光源的偏轉情形，此演算法可以得出每個準直 LED 模組單元角度變動對於目標平面照度的變動並且與實際的光源在偏移角度於一定的範圍下，兩者之間的誤差可以達到 5%以下。
再來以目標平面的照度變動做為神經網路的學習對像，並且提取訓練資料的特徵使得準確率能夠提升，最終成效可以達到 75%以上，以此訓練所分類的角度變動，可以做為調整組裝誤差的依據，藉以提昇每個準直 LED 模組單元的組裝定位精度，進而提昇整個 UV-LED 平行曝光機之光源模組的光學品質。 |
| 摘要(英) |
This thesis is based on the ideal three-dimensional light distribution function surface
generated by symmetrical rotation under the two-dimensional function fitting curve of the light source light distribution, and the function surface formed by superimposing these ideal threedimensional light distribution function surfaces is used as the basis Study the basis function used to simulate the three-dimensional light field of the real physical world. By adjusting the scale mother number and position mother number of the basis function, the square absolute error compared with the optical simulation software can be controlled below 5%.
This thesis takes the UV-LED exposure machine as the basic research object, and uses the three-dimensional light field distribution function to collimate the angle parameter deflection algorithm of the LED module unit. Under the optimal light source spacing of the exposure machine, for different DSR=10 When the uniformity of illuminance on the planes of, 12, 14, 16, and 18 are all greater than 90%, through the symmetry of the light source arrangement and random angle deflection, the illuminance uniformity is distinguished and analyzed, and the deflection of the local light source is added to the uniformity. The range of change is combined with the neural network learning algorithm to distinguish the deflection of each light source. This algorithm can obtain the angle change of each collimated LED module unit to the target plane illuminance change and offset from the actual light source. When the angle is ithin a certain range, the error between the two can reach 4% or less.
Then use the illumination change of the target plane as the learning object of the neural network, and extract the characteristics of the training data to improve the accuracy, and the
final effect can reach more than 75%. The angle change classified by the training can be used
as The basis for adjusting the assembly error is to improve the assembly and positioning
accuracy of each collimated LED module unit, thereby improving the optical quality of the light
source module of the entire UV-LED parallel exposure machine. |
| 關鍵字(中) |
★ 光場分佈函數
★ 神經網路
★ 多光源照度分佈 |
關鍵字(英) |
★ light field distribution function
★ neural network
★ multi-light source illuminance distribution |
| 論文目次 |
摘要........................................................................................................................................I
ABSTRACT ......................................................................................................................... II
致謝..................................................................................................................................... III
目錄.....................................................................................................................................IV
表目錄................................................................................................................................VII
圖目錄.................................................................................................................................IX
第一章、緒論 ....................................................................................................................... 1
1-1 研究背景......................................................................................................................... 1
1-2 研究動機與目的 ............................................................................................................. 5
1-3 文獻回顧......................................................................................................................... 6
1-3-1 光場函數模型化 ...................................................................................................... 6
1-3-2 FAST R-CNN 神經網路........................................................................................... 9
1-3-3 非負矩陣分解法…………………………………………………………………….10
1-4 論文架構....................................................................................................................... 11
第二章、基礎理論.............................................................................................................. 13
2-1 光度學........................................................................................................................... 13
2-1-1 立體角.................................................................................................................... 13
2-1-2 光通量.................................................................................................................... 14
2-1-3 光強度.................................................................................................................... 14
2-1-4 光照度.................................................................................................................... 14
V
2-1-5 輝度 ....................................................................................................................... 14
2-2 餘弦四次方定理 ........................................................................................................... 15
2-3 光的餘弦定理............................................................................................................... 16
2-4 照度均勻度................................................................................................................... 16
2-5 倒傳遞神經網路…………………………………………………………………………17
2-6 卷積神經網路……………………………………………………………………………18
第三章、三維光場分佈函數建立與分析........................................................................... 20
3-1 照度分佈模型化 ........................................................................................................... 20
3-1-1 非線性最小平方迴歸法......................................................................................... 20
3-1-2 側向擬合................................................................................................................ 23
3-2 基底函數....................................................................................................................... 25
3-3 平均絕對誤差理論 ....................................................................................................... 29
3-4 多光源疊加的平坦化分佈計算 .................................................................................... 29
第四章、光分佈函數偏轉演算法 ...................................................................................... 32
4-1 研究規劃....................................................................................................................... 32
4-2 單光源角度偏轉光分佈函數........................................................................................ 32
4-2-1 光源偏移設定........................................................................................................ 33
4-2-2 光源旋轉設定........................................................................................................ 37
4-3 多光源系統排列 ........................................................................................................... 39
4-3-1 多光源之光分佈函數演算法................................................................................. 40
4-3-2 光源最佳間距與最佳均勻度................................................................................. 41
VI
第五章、神經網路學習與訓練結果................................................................................... 44
5-1 倒傳遞神經網路訓練.................................................................................................... 44
5-1-1 輸入特徵向量提取..................................................................................................44
5-1-2 訓練過程與訓練結果....................................................................................................46
5-2 卷積神經網路訓練 ....................................................................................................... 49
5-2-1 輸入特徵圖...........................................................................................................49
5-2-2 訓練過程與訓練結果................................................................................................50
5-3 多光源系統的神經網路學習...............................................................................................52
5-3-1 多光源函數光照特徵提取..............................................................................................52
5-3-2 多光源函數照射情形分類.............................................................................................53
5-3-3 訓練過程與結果.................................................................................................55
第六章、結論與未來展望.................................................................................................. 60
6-1 結論............................................................................................................................... 60
6-2 未來展望....................................................................................................................... 61
參考文獻............................................................................................................................. 62 |
| 參考文獻 |
[1] 林一星，臺灣印刷電路板材料產業回顧與展望，2019，工研院產業科技國際策略發展所。
[2] Top 100 G1obal PCB Manufacture and Output Value 檢自，https://evertiq.com/news/50514
[3] 「微影技術持續精進 半導體工業延續飛躍性成長」，DIGITIMES 企劃，檢自https://bit.ly/2lIeKs4
[4] 志聖工業公司網站 ，檢自 https://www.csun.com.tw/
[5] 川寶科技股份有限公司 生 產 的 UV-LED 平 行 光 曝 光 機 ， 檢 自http://www.cbtech.com.tw/tw/product.php?id=79
[6] 「超高壓 UV 氙氣燈」，上能國際有限公司產品資訊，檢自 http://www.uvtech.com.tw/product_651025.html
[7] 郭信宏，2016，「一種應用於類面光源陣列的光場演算技術之研究」，國立中央大學，博士論文。
[8] 徐安永，2017，「一種應用於準直系統光源的光照度分佈演算之研究」，國立中央大學，碩士論文。
[9] 鍾奕晨，2019，「一種應用於 UV-LED 系統光源設計的光分佈演算法之研究」，國立中央大學，碩士論文。
[10] Girshick, R. (2015). Fast R-CNN. In Proceedings of the IEEE international conference on computer vision.
[11] Lee, D. D. & Seung, H. S. (1999). Learning the parts of objects by non-negative matrix factorization.Nature,401(6755), 788-791.
[12] https://ml-cheatsheet.readthedocs.io/en/latest/loss_functions.html
[13] https://ml-cheatsheet.readthedocs.io/en/latest/loss_functions.html
[14] Franklin A. Graybill and Hariharan K. Iyer. 1994. Regression Analysis: Concepts and Applications, Duxbury Pr. United States.
[15] Maron, M. J., & Lopez, R. J. (1982). Numerical analysis: a practical approach (pp. 283-284). New York: Macmillan.
[16] Ito, Kiyoshi. Encyclopedic Dictionary of Mathematics 2nd. MIT Press.
[17] Bukin, A. D. (2007). Fitting function for asymmetric peaks. arXiv preprint arXiv:0711.4449.
[18] Douglas C. Montgomery, Cheryl L. Jennings and Murat Kulahci. 2011. Introduction to
time series analysis and forecasting. John Wiley & Sons. United States.
[19] C. M. Sparrow. 1916. “On spectroscopic resolving power”. The Astrophysical Journal, Vol. 44, pp. 76.
[20] 陳品硯，2019「一種仿真 LED 平行曝光的調控演算法之研究」，國立中央大學，碩士論文。
[21] https://keras.io/about/
[22] Viola and Jones, "Rapid object detection using a boosted cascade of simple features", Computer Vision and Pattern Recognition, 2001 |
| 指導教授 |
陳奇夆(Chi-Feng Chen)
|
審核日期 |
2022-1-20 |
| 推文 |
 facebook  plurk  twitter  funp  google  live  udn  HD  myshare  reddit  netvibes  friend  youpush  delicious  baidu
|
| 網路書籤 |
 Google bookmarks  del.icio.us  hemidemi myshare
|
