【正文】 考慮到每時、每天、每班、每月、每年、每批次等等。過程質量水平越高，所需樣本越小。 Consider hourly, daily, shifts, monthly, annually, lots, and so on. The better your process is performing, the less frequently you will need to sample. 當前產業(yè)標準趨向于小批量多頻率的抽樣。 Current industry standard tends to favor smaller, more frequent samples. 如果采取消除特別起因行動 (穩(wěn)定過程 ) 并且能力被證明 , 100% 監(jiān)視可能被取消(但是要知道客戶的特殊檢查計劃 ) 建立和維護控制限 Setting Up and Maintaining Control Limits ? 用 2025個樣本計算控制限 ,每個樣本大小為 35。 Calculate the control limits with 20 25 samples (., for the X and R chart that would mean 20 25 samples of size 3 5). ? 如果受控進入最后一步。 If process is in control, go to the last step. ? 如果不受控，找出特殊原因 If process is not in control, try to identify special cause. ? 消除特殊原因，重新收集數(shù)據(jù)，重新計算控制限，直到過程受控 Remove special cause, recollect data, recalculate control limits, … until you find the process is in control. ? 在未來的監(jiān)測中不要隨意的改變控制限，除非過程有永久和渴望的改變。 For future monitoring, do not change the limits unless a permanent, desired change has been made to the process. 兩種數(shù)據(jù)類型控制圖 Two General Kinds of Data 屬性控制圖 ATTRIBUTE ? 使用離散 , 可計的數(shù)據(jù) ? Pass/Fail, Good/Bad, Go/NoGo Information 合格 /不合格 , 好 /壞 , 通過 /不通過等信息 ? Can Be Many Characteristics Per Chart 一張圖可以同時描述許多特性 ? Less Expensive, But Less Information 需要較少的資源 ， 所含信息量亦較少 Ex: 1, 2, 3, 4 etc… Good / Bad Machine 1 , 2 , 3 ... 變量控制圖 VARIABLES 使用連續(xù) , 可測量的數(shù)據(jù) Continuous, Measured Data – Cycle Time, Lengths, Diameters, Drops, etc 周期 , 長度 , 直徑 , 體積 , 等等 ? Generally One Characteristic Per Chart 通常每張圖描述一種特性 ? More Expensive, But More Information 需要更多的資源 , 但所含信息量更多 Ex: Weight = Lbs Thickness = inches (1) 螺釘扭矩在每個裝配線傳輸?shù)淖笄敖请x開 Bolt torque on the front left corner of every transmission ing off the assembly line (2) 每個螺釘離開裝配線傳輸?shù)钠骄ぞ? Average bolt torque of every bolt for each transmission ing off the assembly line (3) 每個發(fā)動機所缺的螺釘數(shù) of missing bolts per engine (4) 每個銷售合同的排字數(shù) of typos per sales contract (5) 每月生產缺陷發(fā)動機的數(shù)目 Number of engines with defects in monthly production (6) 每月生產缺陷發(fā)動機的的百分比數(shù) % of defective engines in monthly production (7) 根據(jù)應收帳款 , 收回它的時間 Per accounts receivable, amount of time it takes to close it (8) 每 100個發(fā)動機的缺陷數(shù) Number of engines with defects per 100 built 練習：什么類型的數(shù)據(jù)？ Exercise: What Type of Data? 屬性型變量 連續(xù)型變量 數(shù)據(jù)類型？ 連續(xù)變量 VS 屬性變量 數(shù)據(jù)分組還是單個數(shù)據(jù) ? 缺陷數(shù) VS 缺陷比例 ? GROUPS (Averages) (n1) INDIVIDUAL VALUES (n=1) XBar R XBar S Individuals Moving Range 缺陷數(shù) 缺陷比例 Is The Probability Of A Defect Low ? If You Know How Many Are Bad, Do You Know How Many Are Good? Poisson Distribution Binomial Distribution Individuals Moving Range NO YES YES 樣本大小一定？ YES NO c Chart u Chart 樣本大小一定？ np Chart NO YES p Chart 如何選擇控制圖 Choosing the Correct Control Chart NOTE: XBar S is appropriate for subgroup sizes (n) of 10 控制圖的主要類型 Major Types of Control Charts ? 變量圖 Variables Charts – IMR (個體 individuals) – XBar (平均 average) ? 特性圖 Attribute Charts – NP (有缺陷的數(shù)字 Number defective) – P (有缺陷的比率 Proportion defective) – C (過失數(shù)量 Number of defects) – U (每個單位的過失數(shù)量 Number of defects / unit) 練習 : 選擇什么類型的控制圖 ? Exercise: What Type of Control Chart? (1) 螺釘扭矩在每個裝配線傳輸?shù)淖笄敖请x開 Bolt torque on the front left corner of every transmission ing off the assembly line (2) 每個螺釘離開裝配線傳輸?shù)钠骄ぞ? Average bolt torque of every bolt for each transmission ing off the assembly line (3) 每個發(fā)動機所缺的螺釘數(shù) of missing bolts per engine (4) 每個銷售合同的排字數(shù) of typos per sales contract (5) 每月生產缺陷發(fā)動機的數(shù)目 Number of engines with defects in monthly production (6) 每月生產缺陷發(fā)動機的的百分比數(shù) % of defective engines in monthly production (7) 根據(jù)應收帳款 , 收回它的時間 Per accounts receivable, amount of time it takes to close it (8) 每 100個發(fā)動機的缺陷數(shù) Number of engines with defects per 100 built Variable Control Charts連續(xù)數(shù)據(jù)控制圖 Xbar R Chart平均值和極差圖 (Xbar R 圖 ) 屬性型變量 連續(xù)型變量 數(shù)據(jù)類型？ 連續(xù)變量 VS 屬性變量 數(shù)據(jù)分組還是單個數(shù)據(jù) ? 缺陷數(shù) VS 缺陷比例 ? GROUPS (Averages) (n1) INDIVIDUAL VALUES (n=1) XBar R XBar S Individuals Moving Range 缺陷數(shù) 缺陷比例 Is The Probability Of A Defect Low ? If You Know How Many Are Bad, Do You Know How Many Are Good? Poisson Distribution Binomial Distribution Individuals Moving Range NO YES YES 樣本大小一定？ YES NO c Chart u Chart 樣本大小一定？ np Chart NO YES p Chart 如何選擇控制圖 Choosing the Correct Control Chart NOTE: XBar S is appropriate for subgroup sizes (n) of 10 有效的連續(xù)數(shù)據(jù)控制圖包括 ... A Valid Variable Control Chart Has… ? Data in time or production sequence 以時間或生產順序排序的數(shù)據(jù) – to show stability, timetotime variation 表示穩(wěn)定性，隨時間的波動 ? A measure of central tendency 對居中趨勢的測量 – to portray behavior of process center 描述過程的居中 ? A measure of variability 對離散程度的測量 ? Control limits 控制極限 – to allow separating mon cause from assignable cause 可用來區(qū)分通常原因和特殊原因 (可歸因原因 ) XBarR charts ( XbarR圖 ) ? X Bar Chart: a plot of the sample means over time. Xbar圖： 反映樣本平均值隨時間的變化 ? R Chart: a plot of the range (difference between highest and lowest values) of a sample over time. R圖： 反映樣本的極差 (樣本中最大值和最小值的差 )隨時間的變化 XbarR圖實例 Minitab File: contains measured data for a main shaft .—see column 1(C1) = NC_Lathe. The data is in subgroups of size 3. Minitab文件 : 包含主軸的測量數(shù)據(jù) ， 數(shù)據(jù)見 C1欄 (NC_Lathe). 數(shù)據(jù)子樣為 3. The . specifications are .060 +/ .003. 產品的規(guī)范是 .060 +/ .003. ============= 1. Check stability with a run chart. 用趨勢圖檢驗過程的穩(wěn)定性 2. Check for normality. 檢驗過程是否是正態(tài)分布 3. Using Minitab, create an Xbar and R Chart— what are your observations? 用 Minitab畫出 XbarR圖，你得出什么觀察結論 ? 4. Do the given specifications (specs) “relate” to the Control Limits on the Xbar Chart? If so, how?