在太陽能產業中, 估計太陽能轉換效率是了解太陽能晶碇品質之重要課題。該估計值也會影響太陽能產業製程之配方, 如長晶前原料的配方。已知少數載流子是工廠內可用來量測太陽能電池品質, 進而估計太陽能轉換效率的重要因素。據我們所知, 目前文獻沒有探討這方面的有效公式, 目前國內使用的估計值也有待改善。本研究利用少數載流子相關的資料提出估計太陽能轉換效率之迴歸公式。
其迴歸公式中三個顯著因子為(1) 某晶碇中四個角落的尾端機率平均值, (2) 某晶碇之所有晶棒的少數載流子尾端機率的第四階動差, (3) 同一個晶棒中所有晶片的少數載流子的第四階動差。本文所提出之迴歸公式中三個顯著因子可解釋太陽能轉換效率變異數的88%。

In the solar industry, estimating the solar transferring rate is an important issue because the estimate can be used to understand the quality of a solar crystal ingot. Also, the estimated value greatly influences the manufacture process, such as the material recipes of the crystal ingot in such industry. Minority carrier lifetimes can be used to measure the quality of solar crystal ingot and further to estimate the solar transferring rate. To our knowledge, there does not exist any effectively implementable formula measuring the solar transferring rate as functions of Minority carrier lifetimes.
This research proposes a regression function used to estimate the solar transferring rate as functions of minority carrier’s lifetimes. The proposed three significant factors in the regression function are (1) the average of the tail probability of four corners of a crystal ingot, (2) the fourth moment of the tail probability of all the crystal bars in a crystal ingot, and (3) the fourth moment of carrier lifetimes of all the cells in a same crystal bar. The accuracy of the proposed regression model via the coefficient of determination R-square, the fraction of variances of the response variable (the solar transferring rate) explained by the regression line (via the proposed three factors), is 88%.

致謝i
摘要i
英文摘要ii
目錄iii
表目錄v
圖目錄vii
第1章緒論 1
1.1 研究背景 1
1.2 太陽能電池的製程介紹 4
1.2.1 坩堝噴塗 5
1.2.2 原料充填 6
1.2.3 多晶鑄造長晶 6
1.2.4 晶碇開方 8
1.2.5 晶體評價 9
1.3 影響少數載流子的製程參數 10
1.4 數學符號與名詞定義 13
1.5 研究假設 15
1.6 研究動機與目的 15
1.7 研究流程與架構 16
第2 章文獻探討 17
2.1 太陽能相關之文獻探討 17
2.1.1 少數載流子(Minority Carrier Lifetime) 17
2.1.2 少數載流子的量測技術 18
2.1.3 太陽能轉換效率 19
2.2 統計方法的文獻探討 20
2.2.1 第四階動差 20
2.2.2 相關係數(Correlation) 21
2.2.3 偏相關係數(Partial Correlation) 22
2.2.4 迴歸分析(Regression) 23
2.2.5 逐步迴歸分析(Stepwise Regression) 26
第3 章研究方法 28
3.1 尾端機率(p(L >= C)) 28
3.2 分群 29
3.2.1 幾何位置分群 29
3.2.2 統計特性分群 29
3.3 潛在因子 30
3.4 以逐步迴歸得到參考的顯著因子 31
3.5 以偏相關係數建議出最終的三個顯著因子 31
3.6 本研究所提出的迴歸公式 32
3.7 績效指標 33
3.7.1 訓練組與測試組的驗證 33
3.7.2 本研究建議的因子組合與其它因子組合的比較 34
第4 章研究結果與未來展望 37
4.1 研究結果 37
4.2 未來展望 37
參考文獻 38
A R統計程式碼 41
B 各訓練組的逐步迴歸分析結果(使用統計軟體Minitab) 43
C 14顆晶碇, 11因子的逐步分析結果 52
D 四個因子組合之績效 54

A. Muller, M. Ghosh, R. Sonnenschein, P. Woditsch. (2006) Silicon for Photovoltaic Applications , Materials Science and Engineering B,134, 257-262.

J.A. Giesecke, M.C.Schubert, B.Michl, F.Schindler, W.Warta. (2011) Minority Carrier Lifetime Imaging of Silicon Wafers Calibrated by Quasi-Steady-State Photoluminescence, Solar Energy Materials and Solar Cells , 95, 1011-1018.

Jianwu Zeng and Wei Qiao. (2011) Short-Term Solar Power Prediction Using an RBF Neural Network ,IEEE, 1-8.

J. Isenberg. S. Riepe, S.W. Glunz, W. Warta. (2002) Carrier Density Imaging (CDI):A Spatially Resolved Lifetime Measurement Suitable for In-Line Process-Control, IEEE, 266-269.

Ken Kitabayashi. (1997) Neural Network Based Estimation of Maximum Power Generation from PV Module Using Environment Information, IEEE Transactions on Energy Conversion, 12, No. 3, 241-247.

Michael Bail, Jorg Kentsch, Rolf Brendel, and Max Schulz. (2000) Lifetime Mapping of Si Wafers by an Infrared Camera, IEEE, 99-103.

Mohammad Ziaur Rahman. (2012) Modeling Minority Carrier’s Recombination Lifetime of p-Si Solar Cell ,International Journal of Renewable Energy Research, 2, 1.

Muztoba Ahmad Khan, Saiful Huque and Azim Mohammad. (2013) A Neural Network Model for Estimating Global Solar Radiation on Horizontal Surface, International Conference on Electrical Information and Communication Technology (EICT), 1-4.

Nian Zhang and Pradeep K. Behera. (2011) Solar Radiation Prediction Based on Recurrent Neural Network Trained by Levenberg-Marquardt Backpropagation Learning Algorithm, IEEE , 1-7.

Paras Mandal, Ashraf Ul Haque, Surya T.S. Madhira and Donna I. Al-Hakeem. (2013) Applying Wavelets to Predict Solar PV Output Power Using Generalized Regression Neural Network, IEEE, 1-5.

R. K Ahrenkiel, D. J. Dunlavy, and B. Simonds. (2010) A New Lifetime Diagnostic System For Photovoltaic Materials, IEEE, 000842-000846.

Ronald A. Sinton. (2012) A Technical Comparison of 5 SEMI Standards Documents on Lifetime , SEMI Workshop.

SW-M. T., Schmesiser B. (2009) Omitting Meaningless Digits in Point Estimates : the Probability Guarantee of Leading-Digit Rules, Operations Research, 57, 109-117.

Wang Zhe, Mi Zengquang, Wang Fei and Liu Liqing. (2011) Regression-based Modeling Method for Solar Radiation Exposure, International Conference on Mechatronic Science, Electric Engineering and Computer, 2567-2570.

W.M.W. Mariam and S. Husni. (2006) Influence of Malaysian Climate on the Efficiency of Polycrystalline Solar Cells, First International Power and Energy Coference PECon, 54-57.

Xingjie Ji, Yeyu Zhu, Wanlong Gu, Pan Pan, Yaxing Liu and Guilong Zhang.(2011) Research on The Applicability of Global Solar Radiation Estimation Model in Henan Province, China , IEEE, 115-118.

王珽玉. (2012/04) 淺談太陽能矽晶片之品質檢驗方法(上), 工業雜誌能源/儲能專欄,164-169.

王珽玉. (2012/05) 淺談太陽能矽晶片之品質檢驗方法(下), 工業雜誌能源/儲能專欄,155-160.

鄧應揚. (2011) 太陽能多晶矽晶錠固化生長之熱流場與雜質輸送研究,國立中央大學機械工程學系博士論文.

陳敬智. (2009) 數值模擬比較多晶矽氧雜質分佈與固液界面形狀關係,國立中央大學機械工程學系碩士論文.

蕭文龍. (2009) 多變量分析最佳入門實用書: SPSS+LISREL(第二版), 碁峰資訊.

林明獻. (2007) 矽晶圓半導體材料技術,全華圖書.