《sixsigma六西格玛DAY课件》由会员分享，可在线阅读，更多相关《sixsigma六西格玛DAY课件（62页珍藏版）》请在金锄头文库上搜索。
1、项目生命周期中的质量控制计划 验证测量系统 贯彻过程控制贯彻过程控制 制订审核计划制订审核计划与过程所有人的交接与过程所有人的交接定义测量分析改进控制 关注正确的CTQ 量化问题 Y = f(X) 找到核心变量 明确必须的改变 实现改变 制订制订/修改质量控制计划修改质量控制计划 过程文档管理过程文档管理 过程控制过程控制 确定过程能力sixsigma六西格玛_DAYQuality Planning质量控制计划sixsigma六西格玛_DAY质量控制计划：学习目的定义质量控制计划和内容 明确质量控制计划是书面文件 说明过程控制计划的必要性 描述监督和审核的必要性和主要考虑的问题回顾过程管理图s
2、ixsigma六西格玛_DAY什么是质量控制计划？质量控制计划是一种书面文件，目的是为了保证每一产品的特性或过程的要求保持一致质量控制计划包括：过程的书面文件和标准 遵循的过程 容差和其它规范过程控制 需要监督和审核的项目 对过程失控的响应计划ISO 9000对建立和维护质量控制计划有严格的要求 对于非生产领域，其它过程标准会提供更好的建立和维护质量控制计划结构sixsigma六西格玛_DAY质量控制计划质量控制计划质量控制计划质量控制计划 - - 书面文件书面文件书面文件书面文件n描述过程的流程描述过程的流程过程的流程流程展开：显示每个过程步骤的责任人n描述标准的操作程序描述标准的操作程序明
3、确：清晰准确地描述采取的行动和时间、地点说明浅显易懂，没有经过系统培训的人员也能很好执行描述怎么预防产品或过程偏差-包括原因和结果的关系提供容差和其它规范提供清晰、明确的指导n检验质量控制计划的有益提示检验质量控制计划的有益提示检验程序检验包含的所有步骤强调程序的重要性提供更新的方法sixsigma六西格玛_DAY质量控制计划-过程控制 监测监测持续测量过程的偏差和持续测量过程的偏差和/或能力或能力在非随机原因偏差和在非随机原因偏差和/或能力不佳时采取措施或能力不佳时采取措施.频率视过程能力、成本、资源等而定 由与系统最接近的人员进行 保持过程受控需要 审核审核对整个过程全面的回顾保证控制存在
4、和有效当前的文件和标准与程序大体一致验证测量系统适当的监测和响应由离系统较远的人员进行过程控制的关键是.sixsigma六西格玛_DAY过程控制-监测1. 为何监测？为何监测？因为内部的查证不够2. 需要监测什么需要监测什么 监测可以在引起客户不满意之前帮助发现和矫正偏差3. 需要收集多少数据？需要收集多少数据？4. 怎样发现过程偏差和能力的变化？怎样发现过程偏差和能力的变化？5. 发现变化该怎么办？发现变化该怎么办？6. 如果过程受控且有能力，客户仍然满意吗？如果过程受控且有能力，客户仍然满意吗？关键问题关键问题sixsigma六西格玛_DAY监测-充分的评估Importance ofReq
5、uirement需求的重要性需求的重要性 Production / Process Method生产生产/过程的方法过程的方法ProcessCapability工程能力工程能力Method of Primary Control主要控制方法主要控制方法Appraisal评估评估Computer NumericControl/ Direct NC电脑数控/网络Tape Control NC MillingPreset Tooling网络预先控制SPC Monitor SPC监测 1 of 50Hand Bench手工(Hand polishing, fitting手工抛光设备)Operator C
6、ontrolled操作者控制Word Processing文字处理Experienced 经验Reviewer校对Training培训Operator Coding操作者代码Drop-down 分解list记录 training培训Sampling PlanInspection抽样检验SPC MonitorSPC监测 1 of 10Frequent Inspection经常检查High高Low低High高 (Financial/Legal Docs财务/法律文件)Medium中 (Past DueRoot Cause Codes过期订单主要原因代码)ZST = 2.5ZST = 6.0ZST
7、= 2.0ZST 4.0sixsigma六西格玛_DAY标准化和文档化标准化和文档化监测过程监测过程 过程管理过程管理图是帮助概略管理过程的流程和矩阵过程管理图是帮助概略管理过程的流程和矩阵 过程步骤是什么 何人何时进行哪里可以找到更详细的工作指导何处收集过程和产品的数据谁收集数据怎样测量和记录结果何时收集数据 (频率)谁根据数据采取行动何处可以找到解决问题的程序在过程失效时采取什么行动响应计划响应计划sixsigma六西格玛_DAY 控制计划的类型 1. 设备控制计划2. 电脑/录音带/模板/工具控制3. 连续数据控制图SPC4. 特性控制计划5. 过程控制计划6. 装配线控制计划7. 机器
8、/测量中心控制计划 sixsigma六西格玛_DAYStatistical Process Control统计过程控制sixsigma六西格玛_DAY统计过程控制: 学习目的描述两种类型的控制图描述两种类型的控制图 学习使用连续和逻辑数据的控制图学习使用连续和逻辑数据的控制图 根据数据类型为给定的过程选择适当的控制图根据数据类型为给定的过程选择适当的控制图 定义控制限定义控制限,明确其与规范限的区别明确其与规范限的区别 说明说明SPC中的决策错误中的决策错误说明判定过程是否失控的规则说明判定过程是否失控的规则(MiniTab规则及规则及West Electric规则规则) 描述和练习在描述和练
9、习在Minitab中画中画IM, U, C, P和和NP图图sixsigma六西格玛_DAY两种类型的控制图使用测量的数值 周期, 长度, 直径, 体积, 等等通常每张图描述一种特性 需要更多的资源, 但所含信息量更多 n合格/不合格, 好/坏, 通过/不通过等信息n一张图可以同时描述许多特性 n需要较少的资源，所含信息量亦较少连续数据控制图连续数据控制图逻辑数据控制图逻辑数据控制图sixsigma六西格玛_DAY选择适当的控制图Variable orAttribute Data?连续连续/逻辑数据？逻辑数据？Defects or % Defective缺陷数/缺陷百比Constant Lot
10、 Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百比cnpHigh or Low Volume数据量大小Low少High多Individuals & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYVariable Control Charts连续数据控制图X-bar R Chart平均值和极差图 (Xbar - R 图)sixsigma六西格玛_DAYSelecting the Appropriate Control Chart 选择
11、适当的控制图Variable orAttribute Data?连续连续/逻辑数据？逻辑数据？Defects or % Defective缺陷数/缺陷百比Constant Lot Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百分比cnpHigh or Low Volume数据量大小Low少High多Individuals & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYA Valid Variable Control Chart
12、Has 有效的连续数据控制图包括.Data in time or production sequence 以时间或生产顺序排序的数据to show stability, time-to-time variation 表示稳定性，随时间的波动A measure of central tendency 对居中趋势的测量to portray behavior of process center 描述过程的居中A measure of variability 对离散程度的测量Control limits 控制极限to allow separating common cause from assigna
13、ble cause 可用来区分通常原因和特殊原因(可归因原因) X-Bar-R charts ( Xbar-R图图)X Bar Chart: a plot of the sample means over time. Xbar图：图：反映样本平均值随时间的变化R Chart: a plot of the range (difference between highest and lowest values) of a sample over time. R图：图：反映样本的极差(样本中最大值和最小值的差)随时间的变化sixsigma六西格玛_DAYX Bar & R Chart Example
14、 Xbar-R图实例 Minitab File: Xbar_r.mtw contains measured data for a main shaft O.D.see column 1(C1) = NC_Lathe. The data is in subgroups of size 3.Minitab文件文件: Xbar_r.mtw 包含主轴的测量数据 ，数据见C1栏(NC_Lathe). 数据子样为3.The O.D. specifications are .060 +/- .003. 产品的规范是.060 +/- .003.=1. Check stability with a run ch
15、art. 用趋势图检验过程的稳定性2. Check for normality. 检验过程是否是正态分布3. Using Minitab, create an Xbar and R Chart what are your observations? 用Minitab画出Xbar-R图，你得出什么观察结论?4. Do the given specifications (specs) “relate” to the Control Limits on the Xbar Chart? If so, how? 给出的产品规范与Xbar图的控制极限相关吗？如果是的话，如何相关？5. How does P
16、rocess Control “relate” to Process Capability? 过程控制如何与过程能力相关？sixsigma六西格玛_DAYX Bar & R Chart Example (cont.) Xbar-R图实例 (续)MINITAB FILE: Xbar_r. mtw MINITAB文件: Xbar_r. mtwsixsigma六西格玛_DAYX Bar & R Chart - Output Xbar-R图结果1. Double Click “C1.” 双击 “C1.”2. Type in a 3 for Subgroup size. 子样大小为3.3. Click
17、“OK.” 点击“OK.”Note that 3.0 SL denotes a 3 sigma limit = Control Limit注意3.0 SL 表示控制极限=3sigma水平 Do not confuse this with specification limits.不要将控制极限与规范极限混淆sixsigma六西格玛_DAYControl Limits vs. Specification Limits控制极限和规范极限The control limits are for averages, not individual values.控制极限是根据平均值计算得出的 Most sp
18、ecifications are for individual values.大多数规范是关于个体数值的USLUCLLCLLSLsixsigma六西格玛_DAYTypes of Errors in Control Charts 控制图中的错误类型 3s-level Control Limits 3s-水平控制极限Created by Shewhart to minimize two types of mistakes 由Shewhart创立，使两种类型的错误最小化Placed empirically because they minimize the two types of mistakes
19、 经验证明，使两种类型的错误最小化Are not probability limits 不是可能性极限Two types of Mistakes:Calling a special cause of variation a common cause of variation (Missing a chance to identify a change in the process) 将特殊原因引起的偏差误作为通常原因引起的偏差 Calling a common cause of variation a special cause of variation (Interfering with a
20、 stable process, wasting resources looking for special causes of variation that do not exist. 将通常原因引起的偏差误作为特殊原因引起的偏差Control limits vs. Specification limits 控制极限和规范极限控制极限和规范极限 Control limits are determined by process capability. 控制极限是由过程能力决定的 Specification limits are determined by customer needs 规范极限
21、是由客户的要求决定的 sixsigma六西格玛_DAYMinitab Rules Minitab规则1. One point beyond zone A.(Western Electric) 一个点在A区外(Western Electric规则 )2. Nine points in a row in zone C or beyond. (All on one side.) (Western Electric) 连续9个点在C区或以外 (在控制图同一边)(Western Electric规则 )3. Six points in a row, all increasing or decreasin
22、g. 连续6个点同时上升或下降 4. Fourteen points in a row, alternating up and down. 连续14个点交替上升和下降 5. Two out of three points in a row in zone A or beyond. (Western Electric) 连续3个点中有2个在A区或以外(Western Electric规则)6. Four out of five points in a row in zone B or beyond. (Western Electric) 连续5个点中有4个在B区或以外(Western Elect
23、ric规则)7. Fifteen points in a row in zone C, above or below center. 连续15个点在中心线两侧的C区内 8. Eight points in a row beyond zone C, above or below center. 连续8个点在中心线两侧的C区外 -3s-2s+2s+3s-1s+1sC C B B A A sixsigma六西格玛_DAYVariable Control Charts：I & MR Chart连续数据控制图：I-MR图sixsigma六西格玛_DAYSelecting the Appropriate
24、Control Chart 选择适当的控制图Variable orAttribute Data?连续连续/逻辑数据？逻辑数据？Defects or % Defective缺陷数/缺陷百比Constant Lot Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百比cnpHigh or Low Volume数据量大小Low少High多Individuals & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYIndividuals & Mo
25、ving Range Charts I-MR图More useful in low volume, intermittent operations 在数据量较少，间歇性操作时更有用Similar to X bar & R Charts, Except 除以下几点外，与Xbar-R图相似Single Values, Not Subgroups 单个数值，不是子样平均值Range Values Must Be Artificially Constructed 极差值需要人工计算Somewhat “Noisier” Because Of Loss Of “Damping” 由于使用个体数值，与Xba
26、r-R图比较更易受干扰 IM Charts IM图图Individuals Chart: a plot of the individual values over time. 个体图个体图 (I图图)：反映个体数值随时间的变化Moving Range Chart: a plot of the moving range (for two samples |Xi - Xi-1|) over time. 移动极差图移动极差图 (MR图图)：反映两个连续样本的移动极差随时间的变化 sixsigma六西格玛_DAYBuilding an Individuals and Moving Range Char
27、t建立I-MR图Individual Data个体数据Moving Range移动极差 55 N/A 56 ABS(55-56) = 1 59 ABS(56-59) = 3 55 ABS(59-55) = 4605958575655Individuals个体个体43210Moving Range移动极差移动极差sixsigma六西格玛_DAYExample:Individuals & Moving Range Chart I-MR图实例: Data from a shaft diameter turning operation are entered on the control chart
28、form on the next page for 25 consecutive pieces of product, in production sequence. 数据是按照生产次序排列的25个连续的转轴产品的直径The data is in Minitab File: Imr.mtw, column shaft_OD. Using Minitab, create the I-MR chart. 数据在Minitab文件文件: Imr.mtw的shaft_OD.栏中，用Minitab画I-MR图Analyze your results. Are there out-of-control i
29、ndications? List the indications, if any, by type and by plot point numbers. 分析结果，过程是否有失控的征兆?根据征兆类型及其编号，列出失控征兆 What is happening in the process? 过程出现了什么情况?sixsigma六西格玛_DAYExample:Individuals & Moving Range Chart I-MR图实例:MINITAB FILE: Imr.mtwMINITAB文件: Imr.mtwsixsigma六西格玛_DAYInput:Individuals & Movin
30、g Range Chart I-MR图输入:1. Double click on “Shaft_OD.” 双击“Shaft_OD.”2. Click “Tests.” 点击 “Tests.”3. Click on “Perform all eight tests.” 点击“ Perform all eight tests. ”sixsigma六西格玛_DAYOutput:Individuals & Moving Range Chart I-MR图结果:sixsigma六西格玛_DAYTake AwaysVariable Control Charts 连续数据控制图小结Variable cont
31、rol charts can be used with continuous data to tell when a process is: 连续数据控制图可以用来区分过程状态:experiencing only common cause variation and working at its intended best 过程只包含通常原因引起的偏差，处于受控状态 when the process is disturbed and needs corrective action 过程受到干扰，需要采取纠正行动控制图:time ordered plot of data 描绘数据随时间的变化re
32、flect the expected range of variation of the data 反映所期望的数据波动的范围 identifies when a special cause appears to be influencing the data 识别何时特殊原因出现，影响数据分布 X-Bar & R charts are used for plotting means and ranges of subgroups over time. Xbar-R图用来描述子样的平均值和极差随时间的变化 I & MR charts are used for plotting individu
33、al values and moving ranges over time. I-MR图用来描述个体的数值和移动极差随时间的变化 sixsigma六西格玛_DAYTake AwaysVariable Control Charts 连续数据控制图小结Control limits are typically calculated as 3 standard deviations away from the mean of the process. 控制极限一般是按过程中心值+/- 3个标准偏差计算出来的 Control limits and specification limits are not
34、 the same. 控制极限和规范极限是不一样的 Control limits are calculated from the sample data; they are internal to the process 控制极限是根据样本数据计算得出的; 是过程的内部特征 Specification limits are determined by your performance standard; they are external to the process 规范极限是由执行的标准决定的; 是过程的外部特征 Know when a process is out of control:
35、 Western Electric Rules. 知道过程何时失控: Western Electric规则 Control charts are only as good as the actions that you take to keep the process under control. 控制图和采取的纠正行动共同使过程保持受控 sixsigma六西格玛_DAYAttribute Control Charts逻辑数据控制图sixsigma六西格玛_DAYSelecting the Appropriate Control Chart 选择适当的控制图Attribute Data逻辑数据
36、逻辑数据Variable or Attribute Data?连续连续/逻辑数据？逻辑数据？Defects or % Defective缺陷数/缺陷百比Constant Lot Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百比cnpHigh or Low Volume数据量大小Low少High多Individuals & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYImportant Definitions 重要定义A Defec
37、t (缺陷缺陷)A single characteristic that does not meet requirements 不满足要求的单一特性A Defective (缺陷率缺陷率)A unit that contains one or more DEFECTS 包含单个或多个缺陷缺陷的单位Attribute Charts Can Consider Either CaseDepending On The Chart Type Chosen根据所选择的控制图类型，逻辑数据控制图可以考虑两者之一的情根据所选择的控制图类型，逻辑数据控制图可以考虑两者之一的情形形sixsigma六西格玛_DAY
38、Classification of Attribute Chart Types 逻辑数据控制图的分类cun ppConstantLot / Unit Size样本数不变VariableLot / Unit Size样本数变化Defects缺陷数Poisson泊松分布Binomial两项式分布Defective缺陷率sixsigma六西格玛_DAYAttribute Control Chart逻辑数据控制图C-Chart (C图)sixsigma六西格玛_DAYC-Chart (C图)Chart for defects per unit (subgroup) 描述每单位或子群所包含的缺陷的控制图
39、Based on Poisson distribution 根据泊松分布而来High probability of finding defect of some type. Large samples are needed if defect probabilities are low. 发现某些缺陷的可能性高；如果缺陷概率低的话，需要大样本。Lower probability of a defect of a given type 特定类型缺陷发生的可能性低Works best on complex unit of product 在产品单元复杂时效果最佳Constant subgroup/
40、lot size 子样大小为常数sixsigma六西格玛_DAYC-Chart ExampleC图实例Attribute Data逻辑数据Manufacturing data indicates that a significant loss occurs from welding nonconformances on part A detected at NDT. The data on the number and general type of nonconformity for each part tested is maintained by serial number in the
41、 NDT log books. 生产数据表明部件A的焊接不合格造成了可观的损失，测试的每个部件不合格的数目和类型都有相应的记录。 To determine the current performance of the welding process we will plot the number of nonconformities per subgroup of two parts on a C-Chart using the data from the log book: 为了确定目前焊接过程的表现，根据记录的数据用C图画出每一子样(2个部件)的不合格数 Date日期日期 6/16/26/
42、36/46/56/86/9Number of 4 2 5 6 10 5 6Nonconf. 3 2 8 5 76per 7 4 9 9 Subgroup每一子样的每一子样的不合格数不合格数 7 5 7 4 5 6 6 7Example Part I.实例第一部分：nThe , column weld_I, contains the data given above. 以上数据在文件C_chart中的weld_I列1.Using Minitab, create a C-Chart. 用Minitab画出C图2.Is the high level of nonconformance we are
43、experiencing due to an assignable cause or random variation? 目前的高水平的不合格品是由可归因原因还是随机原因引起的偏差3.What are some actions for consideration to reduce the level of nonconformances generated by this process? 为了降低过程的不合格水平，需要考虑采取什么措施？4. What is the Process Capability? 过程能力怎样 sixsigma六西格玛_DAYC-Chart Example - Pa
44、rt I C图实例第一部分MINITAB FILE: C chart.mtwMINITAB文件: C chart.mtwsixsigma六西格玛_DAYC-Chart Example - Input C图实例 - 输入sixsigma六西格玛_DAYAttribute Control Charts逻辑数据控制图U-Chart (U图)sixsigma六西格玛_DAYSelecting the Appropriate Control Chart 选择适当的控制图Attribute Data逻辑数逻辑数据据Variable or Attribute Data?连续连续/逻辑数据？逻辑数据？Defe
45、cts or % Defective缺陷数/缺陷百比Constant Lot Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百比cnpHigh or Low Volume数据量大小Low少High多Individuals & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYU-Chart (U图)U-Chart - defects per unit, variable lot (subgroup) size U图描述子样数变化时每单元的
46、缺陷数Same logic as C-Chart, except variable lot (subgroup) size (n) 除了子样数n可变外，原理与C图相同sixsigma六西格玛_DAYU-Chart Example U图实例In .mtw, column “errors” contains time order data of customer parts order defects found each day. A defect is defined to be inaccurate information found on a parts order requisition
47、. Both the number of defects and the daily number of orders are recorded. 在文件U_chart.mtw的“errors”列中记录了每天发现的客户零件订单缺陷的数据. 缺陷定义为在零件订货单上发现的不准确信息，缺陷数和每天的订单数都有记录。 Using Minitab, construct a U-Chart of the data. 用数据在Minitab中画出U图Using the Six Sigma Product Report obtain the Process Capability data for this
48、parts ordering operation. 用Six Sigma Product Report 得出零件订货操作过程的过程能力数据What are your observations?观察结果如何？sixsigma六西格玛_DAYU-Chart Example：Minitab Menu CommandsU图实例： Minitab 命令菜单MINITAB FILE: U_Chart.mtwMINITAB文件: U_Chart.mtwsixsigma六西格玛_DAYU-Chart Minitab Input & Output U图在Minitab中的输入和输出sixsigma六西格玛_DA
49、YAttribute Control Charts逻辑数据控制图P-Chart (P图)sixsigma六西格玛_DAYSelecting the Appropriate Control Chart选择适当的控制图P-Chart (P图图) Variable or Attribute Data?连续连续/逻辑数据？逻辑数据？Defects or % Defective缺陷数/缺陷百比Constant Lot Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百比cnpHigh or Low Volume数据量大小Low少High多Individuals
50、 & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYCommands 命令MINITAB FILE: P chart.mtw (MINITAB文件: P chart.mtw)In .mtw, column “voids” contains data for number of parts containing coating voids found at inspection after a coating operation.在文件在文件P_chart.mtw中的中的“void
51、s”列记录了在涂层操作后的检测中发现的包含无效涂层的零件数。列记录了在涂层操作后的检测中发现的包含无效涂层的零件数。 Using Minitab, create a P-Chart of the data.根据数据在根据数据在Minitab中画出中画出P图图What are your observations? What is needed to estimate the Process Capability?观察结论是什么？为估计过程能力，还需要什么信息？观察结论是什么？为估计过程能力，还需要什么信息？sixsigma六西格玛_DAYP-Chart Minitab Input & Outpu
52、t Minitab中P图的输入和输出Out of controls: determine cause and adjust失控，确定原因并调整Nonconstant control limits due to variable subgroup size子样变化引起的控制限不恒定Very good, determine cause. 非常好，确定原因sixsigma六西格玛_DAYAttribute Control Charts逻辑数据控制图NP-Chart (NP图)sixsigma六西格玛_DAYSelecting the Appropriate Control Chart 选择适当的控制
53、图NP-Chart (NP图图) Variable or Attribute Data?连续连续/逻辑数据？逻辑数据？Defects or % Defective缺陷数/缺陷百比Constant Lot Size?样本数不变？upNo否Yes是Defects or Defective缺陷数/缺陷百比cnpHigh or Low Volume数据量大小Low少High多Individuals & Moving Range个体和移动极差图X-Bar & RangeXbar-R图Variable连续数据Attribute逻辑数据sixsigma六西格玛_DAYNP-Chart (NP图)NP-cha
54、rt: number conforming in subgroup NP图：子样大小一致Same logic as the p-chart, except constant lot size (n) 除子样大小n恒定外，原理同P图sixsigma六西格玛_DAYNP-Chart Example NP图实例In file Np_chart.mtw, column “switches” contains inspection data from 25 consecutive lots of electrical switches. 在文件Np_chart.mtw中的“switches”列中记录了2
55、5组连续 的电器开关子样的测试数据 。 The lot size is constant at 100 switches per lot. 子样大小为每组恒定100个电器开关。Using Minitab, create an Np-Chart of the data. 根据数据在Minitab中画出NP图。What are your observations? 观察结论是什么？What is the Process Capability? What do you need to know to answer this? 过程能力怎样？为了回答这一问题，需要做些什么？sixsigma六西格玛_D
56、AYNP-Chart Example：Minitab Menu CommandsNP图实例： Minitab命令菜单MINITAB FILE: Np chart.mtwMINITAB文件: Np chart.mtwsixsigma六西格玛_DAYNP-Chart Minitab Input & Output NP图的Minitab 输入和输出sixsigma六西格玛_DAYAttribute Chart Subgroup Size 逻辑数据控制图的子样大小Rule of Thumb:拇指规则Select a Subgroup Size that will Provide an Average
57、Defect/Defective Count of Approximately 选择的子样大小须能提供大致平均的缺陷数/缺陷率To Make UCL & LCL Nearly Symmetrical Around the Mean为保证UCL和LCL大致关于平均值对称C, U, NP 5.0For NP-charts, to select the appropriate sample size such that 95% of the subgroups will have at least one defective, use the relationship 对NP图，为选择适当的样本大小
58、以使95%的子样会包含至少一个缺陷，可使用以下关系式sixsigma六西格玛_DAYSummary of Attribute Charts 逻辑数据控制图小结Useful when variable data not available 当连续数据无法得到是很有用Use count/classification datapass/fail, good/bad 使用诸如合格/不合格、好/坏等记数/分类数据Same general rules for interpretation as variable charts 一些通用规则与连续数据控制图相同Useful as end-to-end ove
59、rview; use variable charts for further study of problems对分析概况很有用，使用连续数据控制图对问题可做进一步研究Can use data gathered for other purposes 可使用为其它目的收集的数据Generally less expensive to administer, but tell you less 通常需要较少资源管理，但得到的信息亦较少sixsigma六西格玛_DAYSummary of Attribute Charts 逻辑数据控制图小结 Shortcomings 缺点：缺点：Including t
60、oo many variables makes interpretation difficult 包含太多变量，使解释较为困难Must fit the parameters you are evaluating to theoretical distribution Poisson (C, U-Charts), Binomial (P, NP-Charts) 必须满足所评估的参数的理论分布(C图U图泊松分布、P图NP图二项式分布)Need to evaluate whether constant/non-constant lot size will help you with root cau
61、se analysis 须评估子样大小是否恒定对根本原因分析的影响Sensitivity is dependent on magnitude of defect level敏感性取决于数量和缺陷水平sixsigma六西格玛_DAYTake AwaysAttribute Control Charts 逻辑数据控制图小结Attribute control charts are used to monitor the level of nonconformance of a process. 逻辑数据控制图用来监控过程不合格品的水平Select the appropriate attribute control chart based upon 根据以下两项选择适当的逻辑数据控制图constant vs. variable lot size子样大小是否恒定defects vs. defectives缺陷数/缺陷率Defect (缺陷)A single characteristic that does not meet requirements 不满足要求的单一特性Defective (缺陷率)A unit that contains one or more DEFECTS 包含单个或多个缺陷缺陷的单位sixsigma六西格玛_DAY
