【正文】 29 42 40 35 38 – B B A A B A 2023/2/7 71 Dr. Weihua Zhou Zhejiang University 趨向檢驗 Run Test /( ) 1221()3UDNErNEr????A/B預 期 趨 向 數(shù) ：/1416 2990UDNN??????A/B趨 向 的 標 準 偏 差 ：2023/2/7 72 Dr. Weihua Zhou Zhejiang University N觀察的次數(shù) 當 z在 【 2， +2】 之間，可以認定工序是正常的。 趨向檢驗 樣本 均值 A/B U/D 樣本 均值 A/B U/D 1 B 11 B D 2 B U 12 A U 3 B D 13 B D 4 A U 14 A U 5 B D 15 A D 6 A U 16 A U 7 A D 17 A D 8 A U 18 B D 9 B D 19 B U 10 B U 20 A U rA/B=10 rU/D=17 2023/2/7 73 Dr. Weihua Zhou Zhejiang University 趨向檢驗 131711101331202)(111220)(////????????????DUBADUBAzzrErE－預期趨向數(shù)：902920234120//???????DUBA??趨向的標準偏差：2023/2/7 74 Dr. Weihua Zhou Zhejiang University 非隨機變化 NonRandom Variation ? 管理人員需有對應的計劃來調(diào)查原因 Managers should have response plans to investigate cause ? 可能是錯誤預警（ I型錯誤） May be false alarm (Type I error) ? 可能是可識別的變化 May be assignable variation 2023/2/7 75 Dr. Weihua Zhou Zhejiang University 統(tǒng)計過程控制錯誤 SPC Errors ? I型錯誤 Type I error ( α ) – False alarm – 推斷一個實際處于統(tǒng)計過程控制狀態(tài)的過程不在控制狀態(tài)。 Concluding a process is not in control when it actually is ? II型錯誤 Type II error ( β ) – Missed signal – 推斷一個不在統(tǒng)計過程控制狀態(tài)的過程處于控制狀態(tài) Concluding a process is in control when it is not 2023/2/7 76 Dr. Weihua Zhou Zhejiang University 類型 I和類型 II Type I and Type II Errors 控制中 In control 失控 Out of control 控制中 In control 無錯誤 No Error Type I error (生產(chǎn)者風險producers risk) 失控 Out of control Type II Error (消費者風險consumers risk) 無錯誤 No Error 2023/2/7 77 Dr. Weihua Zhou Zhejiang University I型錯誤 Type I Error 均值 Mean LCL UCL ?/2 ?/2 ? ? I型錯誤出現(xiàn)的概率 Probability of Type I error 2023/2/7 78 Dr. Weihua Zhou Zhejiang University 例子 Example ? 經(jīng)過幾次調(diào)查點處于控制范圍之外，一位經(jīng)理開始思考一個 I型的控制誤差概率（ z = ）那么 α是多少？ After several investigations of points outside control limits revealed nothing, a manager began to wonder about the probability of a Type I error for the control limits used ( z =). What is the α risk for z? UCL = 0 + z 1 = LCL = 0 – z 1 = – 2023/2/7 79 Dr. Weihua Zhou Zhejiang University Type I Error 0 – ?/2 ?/2 =1 – α risk = 2 = 2023/2/7 80 Dr. Weihua Zhou Zhejiang University ? 公差或規(guī)范 Tolerances or specifications – 由工程設(shè)計或客戶要求決定的可接受值的范圍 Range of acceptable values established by engineering design or customer requirements ? 過程方差 Process variability – 過程中的自然變化 Natural variability in a process ? 過程能力 Process capability – 工序能力是否與標準匹配 Process variability relative to specification 過程能力 /工序能力 Process Capability 2023/2/7 81 Dr. Weihua Zhou Zhejiang University 過程能力 Process Capability Lower Specification Upper Specification Process variability matches specifications Lower Specification Upper Specification Process variability well within specifications Lower Specification Upper Specification C. 過程能力不足 process variability exceeds specifications 2023/2/7 82 Dr. Weihua Zhou Zhejiang University 過程能力指數(shù) Process Capability Ratio 過程能力指數(shù) Cp = 標準區(qū)間寬度 specification width 過程區(qū)間寬度 process width 質(zhì)量標準上限 Upper specification – 質(zhì)量標準下限 lower specification 6? Cp = ???????? ??? 3XU T Lo r 3LT LXm in=C pk如果過程是中心對稱的，則使用 If the process is centered use Cp 如果過程不是中心對稱的，則使用 If the process is not centered use Cpk 2023/2/7 83 Dr. Weihua Zhou Zhejiang University If the process is not centered 3? 70 100 160 030100100,30100160min3,3min1601001606??????? ????????? ?????????????LS LUS LCLS LUS LCpkp2023/2/7 84 Dr. Weihua Zhou Zhejiang University 能力指數(shù)的局限性 Limitations of Capability Indexes 1. 過程可能是不穩(wěn)定的 Process may not be stable 2. 過程輸出可能不是正態(tài)分布的 Process output may not be normally distributed 3. 過程不是中心對稱的，卻使用了 Cp Process not centered but Cp is used 2023/2/7 85 Dr. Weihua Zhou Zhejiang University 舉例 Example 機器Machine Standard Deviation 標準差 Machine Capability 過程能力 Cp A B C Cp is desirable Cp = 過程基本符合要求 process is barely capable Cp 過程不符合要求 process is not capable ? 工序能力必須大于等于 ，才能符合工序作業(yè)要求。 ? 工序能力越大，產(chǎn)品符合標準的可能性越大。 2023/2/7 86 Dr. Weihua Zhou Zhejiang University 提高過程能力 Improving Process Capability ? 簡化 Simplify ? 標準化 Standardize ? 無誤操作 Mistakeproof ? 升級器械 Upgrade equipment ? 自動化 Automate 2023/2/7 87 Dr. Weihua Zhou Zhejiang University 田口損失函數(shù) Taguchi Loss Function Cost 目標 Target Lower spec Upper spec 傳統(tǒng)損失函數(shù) Traditional cost function Taguchi cost function 2023/2/7 88 Dr. Weihua Zhou Zhejiang University Outline ? Quality and Quality Management ? Statistic Process Control ? Total Quality Management (TQM) ? 6 Sigma 2023/2/7 89 Dr. Weihua Zhou Zhejiang University 全面質(zhì)量管理 Total Quality Management ? SPC： – 以數(shù)理統(tǒng)計方法為基礎(chǔ)的質(zhì)量控制方法 Using statistical approaches to prevent the bad products to occur ? TQM： – 建立在系統(tǒng)管理思想上的組織性的公司文化 ,持續(xù)不斷地改進質(zhì)量，追求用戶滿意，不斷滿足或超出用戶的期望 A philosophy that involves everyone in an anization in a continual effort to improve quality and achieve customer satisfaction 2023/2/7 90 Dr. Weihua Zhou Zhejiang University 一個小故事 ? IBM Canada : – 缺陷率在 % ——當時北美高水平要求 ? 日本供應商 : – 缺陷零部件和其它零部件用塑料袋分開 – 附一封信：我們不知道你們?yōu)槭裁匆?%的缺陷零部件，為方便起見，我們將之分開包裝 . 2023/2/7 91 Dr. Weihua Zhou Zhejiang University 全面質(zhì)量管理的要素 Elements of TQM ? 持續(xù)改進 Continual improvement ? 標桿法（競爭對手） Competitive benchmarking ? 員工授權(quán) Employee empowerment ? 團隊協(xié)作 Team approach ? 基于事實的決定 Decisions based on facts ? 掌握質(zhì)量管理工具 Knowledge of tools ? 供應商質(zhì)量保證 Supplier quality ? 宣傳發(fā)動 Champion ? 源頭質(zhì)量 Quality at the source