[计算机软件及应用]Minitab教材3.ppt
计算机软件及应计算机软件及应用用Minitab教材教材3p2STEP2决定Y特性决定Y特性收集Y特性数据输入MINITAB数据表进行分析结果说明在收集Y特性时要注意层别和分组。各项的数据要按时间顺序做好相应的整理p3STEP3决定Y特性决定Y特性收集Y特性数据输入MINITAB数据表进行分析结果说明将数据输入MINTAB中，或则在EXCEL中都可以。p4STEP4决定Y特性決定Y特性收集Y特性數據輸入MINITAB數據表進行分析結果說明利用MINITABSTATQUALITY TOOLCAPABILITY ANALYSIS (NORMAL)p5STEP5决定Y特性決定Y特性收集Y特性數據輸入MINITAB數據表進行分析結果說明利用MINITAB的各项图形来进行结果说明p6练习样本X1X2X3X4X5199.7098.72100.24101.28101.20299.32100.97100.8799.2498.21399.8999.83101.4899.56100.90499.1599.7199.1799.3098.80599.66100.80101.06101.16100.45697.7498.8299.2498.6498.737101.18100.2499.6299.3399.918101.54100.96100.62100.67100.499101.49100.6799.36100.38102.101097.1698.2697.59100.0999.78p7输入数据注意输入方式 Select: Stat Quality Tools Capabilty Analysis(Normal)p8输入选项输入上下规格界限根据不同的数据输入方式选择分析方法p9选择标准差的估计方法一般选择复合的标准差估计方式p10选项的输入选择是否作正态型转换如果需要计算Cpm则需要输入目标值过程能力表现形式的选择p11以Cpk, Ppk结果的输出103.5102.5101.5100.599.598.597.596.5TargetUSLLSLProcess Capability Analysis for X1 - X5PPM TotalPPM USLPPM USLPPM USLPPM USLPPM USLPPM USLPPM USLPPM USLPPM USLPPM Quality Tools Capabilty Analysis(Weibull)p18填入选项要求韦氏分布的参数估计p19结果图形104102100989694USLLSLProcess Capability Analysis for X1 - X5Calculations Based on Weibull Distribution ModelPPM TotalPPM USLPPM USLPPM basic statisticnormality test 但数据要放到同一个column中，所以必须针对前面的数据进行一下处理p21数据调整进行数据的堆积p22填写选项输入变量输入作为参考的概率记号p23结果输出P-Value: 0.515A-Squared: 0.324Anderson-Darling Normality TestN: 50StDev: 1.12483Average: 99.9056102101100999897.999.99.95.80.50.20.05.01.001Probabilityx1Normal Probability Plotp24结果输出(加标0.5概率)P-Value: 0.515A-Squared: 0.324Anderson-Darling Normality TestN: 50StDev: 1.12483Average: 99.905610210110099989799.90560.50000.999.99.95.80.50.20.05.01.001Probabilityx1Normal Probability Plotp25计量型制程能力分析总结 一般的正态分布使用 Capability Analysis (Normal) 如果是正态分布且其组内和组间差异较大时可用 Capability Analysis (Between/Within) 当非正态分布时则可以使用 Capability Analysis (Weibull)p26Capability Sixpack (Normal) 复合了以下的六个图形 Xbar R 原始数据分布（plot） 直方图 正态分布检定 CPK, PPKp27练习 请以前面的数据来进行相应的Capability Sixpack (Normal)练习 Select: Stat Quality Tools Capabilty Sixpack(Normal)p28输入各项参数输入规格p29选定判异准则选择判异准则p30选择标准差估计方法默认值是复合标准差计算公式p31考虑可选择项如果希望计算Cpm,则输入目标值p32结果输出109876543210101.5100.599.598.5Xbar and R ChartSubgrMean11Mean=99.91UCL=101.1LCL=98.744.53.01.50.0RangeR=2.025UCL=4.282LCL=0109876543210Last 10 Subgroups102.5101.099.598.0Subgroup NumberValues103T 97Capability PlotProcess ToleranceIIIIIIIIISpecificationsWithinOverall10210098Normal Prob Plot10210098Capability HistogramWithinStDev:Cp:Cpk:0.8706231.151.11OverallStDev:Pp:Ppk:Cpm:1.130590.880.860.89Process Capability Sixpack for X1 - X5p33Capability Sixpack (Between/Within) 复合了以下的六个图形 Xbar R 原始数据分布 直方图 正态分布检定 CPK, PPKp34同前练习及结果103.5101.098.596.0I-MR-R ChartMeanMean=99.91UCL=102.9LCL=96.923210Mov.RangeR=1.124UCL=3.672LCL=01098765432104.53.01.50.0SubgroupRangeR=2.025UCL=4.282LCL=0103T 97Capability PlotProcess ToleranceIIIIIIIIISpecificationsBetween/WithinOverall103.0100.598.0Normal Prob Plot10210098Capability Histogram0.91720.87061.26461.13060.790.770.880.860.89StDevsBetw:Within:Total:Overall:CapabilityCp:Cpk:Pp:Ppk:Cpm:Process Capability Sixpack for X1 - X5p35Capability Sixpack (Weibull) 复合了以下的六个图形 Xbar R 原始数据分布 直方图 正态分布检定 CPK, PPKp36结果输出109876543210101.5100.599.598.5Xbar and R ChartSubgrMeansMean=99.91UCL=101.1LCL=98.744.53.01.50.0RangesR=2.025UCL=4.282LCL=0109876543210Last 10 Subgroups102.5101.099.598.0Subgroup NumberValues103 97Overall (LT)Shape: 102.700Scale: 100.439Pp: 0.74Ppk: 0.64Capability PlotProcess ToleranceSpecificationsIIIIII10210110099989796Weibull Prob Plot10210098Capability HistogramProcess Capability Sixpack for X1 - X5p37二项分布制程能力分析 二项分布只适合用在 好，不好 过，不过 好，坏 不可以用在 0,1,2,3等二项以的选择，此种状况必须使用卜氏分布。p38示例 数据在excel档案中Select: Stat Quality Tools Capabilty Analysis(Binomial)p39填好各项的参数输入样本数输入历史的不良率p40选好控制图的判异准则p41填入选择项可以选择输入图形的表现形式p42结果及输出25201510500.060.050.040.030.020.010.00Sample NumberProportionP=0.01658UCL=0.04367LCL=02520151052.12.01.91.81.71.61.51.41.3Sample Number%Defective4.53.01.50.0Target25020015010076543210%DefectiveSample SizeBinomial Process Capability Report for 不良數Summary StatsCumulative %DefectiveDist of %DefectiveP ChartRate of Defectives(denotes 95% C.I.)Average P:%Defective:Target:PPM Def.:Process Z:0.01658031.6580165802.130(0.0132, 0.0206)(1.32, 2.06)(13168, 20594)(2.042, 2.221)p43卜氏分布制程能力分析 卜分布只适合用在 计数型，有二个以上的选择时 例如可以用在 外观检验，但非关键项部份 0,1,2,3等二项以的选择，此种状况必须使用卜氏分布。p44示例 数据在excel档案中Select: Stat Quality Tools Capabilty Analysis(Poisson)p45填好各项的参数p46结果及输出25201510500.080.070.060.050.040.030.020.010.00Sample NumberSample CountU=0.02342UCL=0.05588LCL=02520151050.0280.0260.0240.0220.020Sample NumberDPU0.0450.0300.0150.000Target2502001501000.040.030.020.01DPUSample SizePoisson Process Capability Report for 不良數_1Summary StatsCumulative DPUDist of DPUU ChartDefect Rate(denotes 95% C.I.)Mean DPU:Min DPU:Max DPU:Targ DPU:0.02341970.0120.040(0.0193011, 0.0281569)p47基础统计p48描述性统计Select: Stat Basic Statistics Display descriptive statisticsp49描述性统计 输出 结果p50平均值检定（Z/t）Select: Stat Basic Statistics 2 Sample tp51比率检定（1P/2P）p52比率检定（1P/2P）输出结果p53标准差的检定（F）p54标准差的检定（F）40302010095% Confidence Intervals for SigmasHouseAge1009080706050403020100Boxplots of Raw DataP-Value : 0.000Test Statistic: 165.617Levenes TestP-Value : 0.000Test Statistic: 0.001F-TestFactor LevelsAgeHouseTest for Equal Variances
