Download presentation
Presentation is loading. Please wait.
2
&SixAdvanced Statistics_DOE11 實驗目的 : &13 DOE 簡介 對 y 影響最大的變數為 何? 如何設定 x 1, x 2, …, x p 使 y 值趨近最佳值? 如何設定 x 1, x 2, …, x p 使 y 值得變異最小? 如何設定 x 1, x 2, …, x p 使不可控制因素 z 1, z 2, …, z p 之影響最小?
3
&SixAdvanced Statistics_DOE12 An Example : Play Golf Objective: Lower score without much practicing. Response Variable: Score (per round) Possible Factors: The type of driver used (oversized or regular-sized) The type of ball used (balata or three-piece) Walking or riding in a golf cart Beverage Type (water or beer) Time (in the morning or afternoon) Weather (cool or hot, windy or calm) The type of golf shoe spike (metal or soft)
4
&SixAdvanced Statistics_DOE13 一般實驗進行方式 Best-guess approach No Good, Guess Again Switching the levels of one (perhaps two) factors for the next test based on the outcome of the current test Good Enough, Stop! On-factor-at-a-time Selecting a baseline starting point Varying each factor over its range with the other factors held constant at the baseline level Interactions ruin everything
5
&SixAdvanced Statistics_DOE14 最佳因子水準組合為? Driver: Regular Mode of travel: Ride Beverage: Water But what if………………… Results of the one-factor-a-time strategy for the golf experiment
6
&SixAdvanced Statistics_DOE15 The two-factor factorial design for the golf experiment (I)
7
&SixAdvanced Statistics_DOE16 The two-factor factorial design for the golf experiment (II) Ball Effect = ? Ball-Driver Interaction Effect = ?
8
&SixAdvanced Statistics_DOE17 Other Designs for the Golf Experiment Four-factor factorial design Three-factor factorial design
9
&SixAdvanced Statistics_DOE18 Other Designs for the Golf Experiment Four-factor fractional factorial design
10
&SixAdvanced Statistics_DOE19 實驗計劃法 (DOE) 在一個或連串的試驗中刻意地改變製程輸入參數值, 以 便觀察並找出影響製程輸出變數之因素. 應用 : 改進製程產出率 降低製程變異, 改善產品品質 降低研發時間 降低總體成本 評估各種可行之設定值 評估各替代原料 確定影響產品特性之因素
11
&SixAdvanced Statistics_DOE110 Example: Optimizing a Process
12
&SixAdvanced Statistics_DOE111 基本原則 複製 (Replication) 隨機化 (Randomization) 區隔化 (Blocking) 增進實驗之精確度 估計自然誤差 中央極限定理 “Averaging out” the effects from uncontrollable variables
13
&SixAdvanced Statistics_DOE112 DOE 之程序 問題之認知與陳述 選擇因子與其水準 選擇反應變數 選擇適當之實驗設計 執行實驗 資料分析 結論與建議 Follow-up run and confirmation test Iterative No more than 25% of available resources should be invested in the first experiment
14
&SixAdvanced Statistics_DOE113 Notes 使用統計以外之專業知識 實驗之設計與分析應愈簡單愈好 實驗之統計分析結果與現實上之差異 成本 技術 時間 實驗通常是遞迴式的 前幾次實驗通常只是學習經驗而已
15
&SixAdvanced Statistics_DOE114 二因子實驗設計 二因子無交互作用
16
&SixAdvanced Statistics_DOE115 二因子有交互作用
17
&SixAdvanced Statistics_DOE116 One-factor at a time 之方法
18
&SixAdvanced Statistics_DOE117
19
&SixAdvanced Statistics_DOE118 二因子實驗設計之模式
20
&SixAdvanced Statistics_DOE119 Data Sheet
21
&SixAdvanced Statistics_DOE120 ANOVA 表 – Two-Factor Factorial
22
&SixAdvanced Statistics_DOE121
23
&SixAdvanced Statistics_DOE122 Example
24
&SixAdvanced Statistics_DOE123
25
&SixAdvanced Statistics_DOE124 決策模式: 因為 F 0 (Primer Types) = 28.63 > F 0.05,2,12 = 3.89 F 0 (Application Methods) = 61.38 > F 0.05,1,12 = 4.75 所以此二因子對黏著力皆有顯著影響。 但 F 0 (Interaction) = 1.5 < F 0.05,2,12 = 3.89 ,所以此二因子的交互作 用對黏著力無明顯之影響。
26
&SixAdvanced Statistics_DOE125 Another Example
27
&SixAdvanced Statistics_DOE126 ANOVA Results Computer Output (Model Adequacy Checking)
28
&SixAdvanced Statistics_DOE127 Multiple Comparisons
29
&SixAdvanced Statistics_DOE128 &6 2 k 因子階層設計 k 個因子,每個因子 2 個水準 (+,-) ,共 2 k 次實驗(當 n = 1 時)。 在因子數不多的狀況下,常用於實驗初期,來了解因 子對反應變數之可能影響。 只能看出因子對反應變數之線性作用 (linear effect) , 無法預估高階曲面作用。
30
&SixAdvanced Statistics_DOE129 The 2 2 Factorial Design
31
&SixAdvanced Statistics_DOE130 The Calculation (I) Effects Contrast
32
&SixAdvanced Statistics_DOE131 Sum of Square Errors The Calculation (II)
33
&SixAdvanced Statistics_DOE132 The AVONA Table
34
&SixAdvanced Statistics_DOE133 The Regression Model Since the effect of AB is not significant, the regression model would be The estimates of these coefficients are Why?
35
&SixAdvanced Statistics_DOE134 The Residuals Residuals at x 1 =1 and x 2 = -1 Computer Output
36
&SixAdvanced Statistics_DOE135 2 3 因子階層設計
37
&SixAdvanced Statistics_DOE136 2 3 因子階層設計 _ 符號表
38
&SixAdvanced Statistics_DOE137 計算 各因子之 Contrast = (Sign) (Treatment Combination ) Contrast A = -(1)+a-b+ab-c+ac-bc+abc Contrast B = -(1)-a+b+ab-c-ac+bc+abc Contrast AB = 各因子之效用 Effect = Contrast / (n 2 k-1 ) 各因子之 Sum of Square = Contrast / (n 2 k ) SS A = {-(1)+a-b+ab-c+ac-bc+abc} 2 / 8n
39
&SixAdvanced Statistics_DOE138 Example for 2 3 Design A: 速度 B: 切割深度 C: 切刀角度
40
&SixAdvanced Statistics_DOE139 計算 _Example 各因子之效用 各因子之 Sum of Square SS A = {-(1)+a-b+ab-c+ac-bc+abc} 2 / 8n = (27) 2 / (8 2) = 45.5625
41
&SixAdvanced Statistics_DOE140 ANOVA 表 _Example
42
&SixAdvanced Statistics_DOE141 Example for 2 4 Design
43
&SixAdvanced Statistics_DOE142 2 4 因子階層設計 _ 符號表 請完成請完成
44
&SixAdvanced Statistics_DOE143 ANOVA 表 _Example
45
&SixAdvanced Statistics_DOE144 AD 交互作用與迴歸函數
46
&SixAdvanced Statistics_DOE145 2 k Design with Center Points 增加預估曲線作用之能力 不破壞設計之平衡性 (Balanced Design) 只需增加少數幾個實驗
47
&SixAdvanced Statistics_DOE146 Example
48
&SixAdvanced Statistics_DOE147 ANOVA 表 _Example
49
&SixAdvanced Statistics_DOE148 2 k 因子實驗之區隔化與混雜化 2 2 factorial design with Blocking
50
&SixAdvanced Statistics_DOE149 The ANOVA
51
&SixAdvanced Statistics_DOE150 Confounding( 混雜化 ) 受限於資源(時間、金錢、人力等),無法再每一個 區隔皆有完整的因子實驗。 Confounding is a design technique for arranging a complete factorial design in blocks, where the block size is smaller than the number of treatment combinations in one replicate. It causes information about certain treatment effects (usually high-order interactions) to be indistinguishable from, or confounded with, blocks.
52
&SixAdvanced Statistics_DOE151 Confounding in 2 blocks Effect AB confounded with block.
53
&SixAdvanced Statistics_DOE152 The 2 3 Design Confounded(I)
54
&SixAdvanced Statistics_DOE153 The 2 3 Design Confounded(II) The Degrees of Freedom
55
&SixAdvanced Statistics_DOE154 The 2 3 Design Confounded(III) Replications => n = 4.
56
&SixAdvanced Statistics_DOE155 The 2 3 Design Confounded(IV) Replications => n = 4.
57
&SixAdvanced Statistics_DOE156 An Example
58
&SixAdvanced Statistics_DOE157 The ANOVA
59
&SixAdvanced Statistics_DOE158 Confounding in 4 blocks ADE and BCE are confounded, in addition, ABCD is also confounded. Why? Suggested Blocking => pp. 298, Table 7-8.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.