系統識別號 | U0002-0107201913572300 |
---|---|
DOI | 10.6846/TKU.2019.00013 |
論文名稱(中文) | Pareto分配產品的壽命績效指標在逐步型I區間設限下之檢定程序的檢定力分析 |
論文名稱(英文) | The power analysis of the testing procedure for the lifetime performance index of products with Pareto distribution under progressive type I interval censoring |
第三語言論文名稱 | |
校院名稱 | 淡江大學 |
系所名稱(中文) | 統計學系應用統計學碩士班 |
系所名稱(英文) | Department of Statistics |
外國學位學校名稱 | |
外國學位學院名稱 | |
外國學位研究所名稱 | |
學年度 | 107 |
學期 | 2 |
出版年 | 108 |
研究生(中文) | 林其儒 |
研究生(英文) | Chi-Ju Lin |
學號 | 606650173 |
學位類別 | 碩士 |
語言別 | 繁體中文 |
第二語言別 | |
口試日期 | 2019-06-14 |
論文頁數 | 77頁 |
口試委員 |
指導教授
-
吳淑妃
委員 - 吳錦全 委員 - 王智立 |
關鍵字(中) |
製程能力指標 逐步型I區間設限 柏拉圖分配 最大概似估計量 拔靴法 檢定程序 |
關鍵字(英) |
process capability index progressive type I interval censoring Pareto distribution maximum likelihood estimator bootstrap testing procedure |
第三語言關鍵字 | |
學科別分類 | |
中文摘要 |
當科技加速進步,人們對於產品的要求也日益嚴苛,不僅廠商希望在限制的時間、成本下,提高品質以追求利潤;消費者更是期望能購買到高壽命、低故障率的產品。在實務上,製程能力指標(process capability indices, PCIs)被廣泛用來評估製程,在雙方的需求下,製程能力能否被有效評估,成為非常重要的關鍵因素。 因此,本研究假設產品壽命服從柏拉圖分配時,在逐步型I區間設限下,計算出壽命績效指標C_L之最大概似估計量並求得其漸近分配。在規格下限L已知的情形下,使用此估計量及兩種拔靴法發展三個檢定程序以評估壽命績效是否達到預定的水準;接著對三種檢定程序計算其模擬檢定力,並進行三種檢定程序之比較分析。最後,利用一個模擬資料與一個實務資料,說明如何利用本研究提出的檢定程序評估產品績效指標是否達到要求水平。 |
英文摘要 |
As the science and technology make progress day by day, people are in the pursuit of more stringent product quality requirements. Not only manufacturers hope to improve quality and pursue profits under limited time and cost; but also consumers are expecting to purchase products with high life expectancy and low failure rate. In practice, process capability indices (PCIs) are widely used to evaluate the capabilities of manufacturing processes. Under the highly demand of both parties, whether the process capability can be effectively evaluated becomes a very important key factor. In this thesis, the lifetime of products is assumed to have Pareto distribution. The maximum likelihood estimator is used to estimate the lifetime performance index based on the progressive type I interval censored sample and its asymptotic distribution is also derived. The MLE and two kinds of Bootstrap methods are developed to construct three testing procedures about the lifetime performance index. The comparisons of power analysis of three methods are done and analyzed. Finally, one simulate example and one practical example are given to illustrate the use of these three testing algorithmic procedure to determine whether the process is capable. |
第三語言摘要 | |
論文目次 |
目錄 目錄 I 表目錄 III 圖目錄 X 第一章 緒論 1 1.1研究動機與目的 1 1.2文獻探討 3 1.2.1製程能力指標之發展 3 1.2.2 設限形式 4 1.3本文架構 6 第二章 壽命績效指標與其估計 7 2.1產品的壽命績效指標 8 2.2壽命績效指標的估計量 11 第三章 壽命績效指標的檢定演算程序與檢定力 16 3.1壽命績效指標的檢定演算程序 16 3.2 樣本數大小之決定 21 3.3檢定力之模擬分析 25 3.4點估計 31 第四章 數值實例示範 32 4.1模擬資料 32 4.2真實資料 36 第五章 結論與未來研究 41 5.1 結論 41 5.2 未來研究 42 參考文獻 43 表目錄 表2.1 壽命績效指標C_L值及對應製程良率Pr 10 附表 1 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60,80,100、觀測次數m=10,20,30、逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,檢定力函數h(c_1)在實際值c_1=0.825(0.025)0.95的數值 45 附表 2 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60,80,100、觀測次數m=10,20,30、逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,檢定力函數h(c_1)在實際值c_1=0.825(0.025)0.95的數值 46 附表 3 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60,80,100、觀測次數m=10,20,30、逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.1下,檢定力函數h(c_1)在實際值 c_1=0.825(0.025)0.95的數值 47 附表 4 當規格下限L=0.05、總觀測時間T=0.5、型II誤差β=0.25,0.2,0.15、觀測次數m=10,20,30、逐步設限移除率p=0,0.05,0.075及實際值c_1=0.825(0.025)0.95時,在顯著水準α=0.01下,所需要的最小樣本數 48 附表 5 當規格下限L=0.05、總觀測時間T=0.5、型II誤差β=0.25,0.2,0.15、觀測次數m=10,20,30、逐步設限移除率p=0,0.05,0.075及實際值c_1=0.825(0.025)0.95時,在顯著水準α=0.05下,所需要的最小樣本數 49 附表 6 當規格下限L=0.05、總觀測時間T=0.5、型II誤差β=0.25,0.2,0.15、觀測次數m=10,20,30、逐步設限移除率p=0,0.05,0.075及實際值c_1=0.825(0.025)0.95時,在顯著水準α=0.1下,所需要的最小樣本數 50 附表 7 當L=0.05、T=0.5、n=60、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.01下,模擬三種方法檢定力的數值 51 附表 8 當L=0.05、T=0.5、n=80、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.01下,模擬三種方法檢定力的數值 52 附表 9 當L=0.05、T=0.5、n=100、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.01下,模擬三種方法檢定力的數值 53 附表 10 當L=0.05、T=0.5、n=60、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.05下,模擬三種方法檢定力的數值 54 附表 11 當L=0.05、T=0.5、n=80、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.05下,模擬三種方法檢定力的數值 55 附表 12 當L=0.05、T=0.5、n=100、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.05下,模擬三種方法檢定力的數值 56 附表 13 當L=0.05、T=0.5、n=60、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.1下,模擬三種方法檢定力的數值 57 附表 14 當L=0.05、T=0.5、n=80、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.1下,模擬三種方法檢定力的數值 58 附表 15 當L=0.05、T=0.5、n=100、m=10,20,30及p=0,0.05,0.075時,在c_0=0.8和α=0.1下,模擬三種方法檢定力的數值 59 附表 16 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 60 附表 17 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=80、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 61 附表 18 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=100、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 62 附表 19 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 63 附表 20 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=80、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 64 附表 21 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=100、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 65 附表 22 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.1下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 66 附表 23 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=80、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.1下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 67 附表 24 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=100、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.1下,模擬三種不同方法的bias在實際值c_1=0.825(0.025)0.95時的數值 68 附表 25 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 69 附表 26 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=80、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 70 附表 27 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=100、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.01下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 71 附表 28 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 72 附表 29 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=80、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 73 附表 30 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=100、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.05下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 74 附表 31 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=60、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.1下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 75 附表 32 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=80、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準 α=0.1下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 76 附表 33 當規格下限L=0.05、總觀測時間T=0.5、觀測樣本數n=100、觀測次數m=10,20,30及逐步設限移除率p=0,0.05,0.075時,在顯著水準α=0.1下,模擬三種不同方法的MSE在實際值c_1=0.825(0.025)0.95時的數值 77 圖目錄 圖1.1 逐步型I區間設限 6 圖3.2.1 當α=0.05、β=0.25和m=20下,不同逐步設限移除率p=0,0.05,0.075時,所需要的最小樣本數n 23 圖3.2.2 當α=0.05、β=0.25和p=0.05下,不同觀察次數m=10,20,30時,所需要的最小樣本數n 23 圖3.2.3 當α=0.05、m=20和p=0.05下,不同型II誤差β=0.25,0.2,0.15時,所需要的最小樣本數n 24 圖3.2.4 當β=0.25、m=20和p=0.05下,不同顯著水準α=0.01,0.05,0.1時,所需要的最小樣本數n 24 圖3.3.1 當α=0.01、n=80及m=20下,不同的逐步設限移除率p=0,0.05,0.075時的檢定力 29 圖3.3.2 當α=0.01、n=80及p=0.05下,不同觀察區間個數m=10,20時的檢定力 29 圖3.3.3 當α=0.01、m=20及p=0.05下,不同樣本數n=60,80時的檢定力 30 圖3.3.4 當n=80、m=20及p=0.05下,不同的顯著水準α=0.01,0.05,0.1時的檢定力 30 |
參考文獻 |
[1] Boyles, R. A. (1991). The Taguchi capability index, Journal of Quality Technology, 23, 17-26. [2] Chan, L. K., Cheng, S. W. and Spiring, F. A. (1988). A new measure of process capability, Journal of Quality Technology, 20(3), 162-175. [3] Efron, B. (1982). The Jackknife, the Bootstrap and other re-sampling plans, CBMS-NSF Regional Conference Series in Applied Mathematics, 38, SIAM, Philadelphia, PA. [4] Gill, M. H. and Gastwirth J. L. (1978). A Sacle-Free Goodness-of-Fit Test for the Exponential Distribution Based on the Gini Statistic, Journal of the Royal Statistical Society, SeriesB(Methodological), 40, 350-357. [5] Hall, P. (1988). Theoretical comparison of bootstrap confidence intervals, Annals of Statistics, 16, p. 927–953. [6] Hong, C. W., Wu, J. W. and Cheng, C. H. (2008). Computational procedure of performance assessment of lifetime index of Pareto lifetime businesses based on confidence interval, Applied Soft Computing , 8, 698–705. [7] Juran, J. M. (1974). Journal quality control handbook(3rd ed), McGraw-Hill ,New York. [8] Kane, V. E. (1986). Process capability indices, Journal of Quality Technology, 18, 41-52. [9] Lawless, J. F. (2003). Statistical Models & Methods for Lifetime Data, (2nded), New York, John Wiley. [10] Lee, H. M., Wu, J. W. and Lei, C. L. (2013). Assessing the lifetime performance index of exponential products with step-stress accelerated life-testing data, IEEE Transactions on Reliability, 62(1), p. 296–304. [11] Lee, W. C., Wu, J. W. and Lei, C. L. (2010). Evaluating the lifetime performance index for the exponential lifetime products, Applied Mathematical Modelling, 34(5), p. 1217–1224. [12] Montgomery, D. C.(1985). Introduction to statistical quality control, John Wiley & Sons, New York. [13] Nigm, A. M., Al-Hussaini, E. K. and Jaheen, Z. F. (2003). Bayesian one-sample prediction of future observations under Pareto distribution, Statistics, Theoretical and Applied Statistics, 37, 527-536. [14] Tong, L. I., Chen, K. T. and Chen, H. T. (2002). Statistical testing for assessing the performance of lifetime index of electronic components with exponential distribution, International Journal of Quality & Reliability Management, 19(7), 812-824. [15] Wu, S. F ., Lu, J. Y. (2017). Computational testing algorithmic procedure of assessment for lifetime performance index of Pareto products under progressive type I interval censoring, Computational Statistics, 32,647–666. |
論文全文使用權限 |
如有問題,歡迎洽詢!
圖書館數位資訊組 (02)2621-5656 轉 2487 或 來信