MINITAB之製程能力分析易騰涂順章1MINITAB之製程能力分析易騰涂順章1製程能力之分類計(jì)量型(基於正態(tài)分佈)計(jì)數(shù)型(基於二項(xiàng)分佈)計(jì)數(shù)型(基於卜氏項(xiàng)分佈)2製程能力之分類計(jì)量型(基於正態(tài)分佈)計(jì)數(shù)型(基於二項(xiàng)分佈)計(jì)MINITAB能力分析的選項(xiàng)(計(jì)量型)CapabilityAnalysis(Normal)CapabilityAnalysis(Between/Within)CapabilityAnalysis(Weibull)CapabilitySixpack(Normal)CapabilitySixpack(Between/Within)CapabilitySixpack(Weibull)3MINITAB能力分析的選項(xiàng)(計(jì)量型)CapabilityCapabilityAnalysis(Normal)該命令會(huì)劃出帶理論正態(tài)曲線的直方圖，這可直觀評(píng)估數(shù)據(jù)的正態(tài)性。輸出報(bào)告中還包含過程能力統(tǒng)計(jì)表，包括子組內(nèi)和總體能力統(tǒng)計(jì)。4CapabilityAnalysis(Normal)該命CapabilityAnalysis(Between/Within)該命令會(huì)劃出帶理論正態(tài)曲線的直方圖，可以直觀評(píng)估數(shù)據(jù)的正態(tài)性。該命令適用於子組間存在較變差的場(chǎng)合。輸出報(bào)告中還包含過程能力統(tǒng)計(jì)表，包括子組間／子組內(nèi)和總體能力統(tǒng)計(jì)。5CapabilityAnalysis(Between/WCapabilityAnalysis(Weibull)該命會(huì)會(huì)劃出帶韋伯曲線的直方圖，這可直觀評(píng)估數(shù)據(jù)是否服從韋伯分布。輸出報(bào)告中還包含總體過程總能力統(tǒng)計(jì)6CapabilityAnalysis(Weibull)該製程能力分析做法決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明7製程能力分析做法決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)STEP1決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明Y特性一般是指客戶所關(guān)心所重視的特性。Y要先能量化，儘量以定量數(shù)據(jù)為主。Y要事先了解其規(guī)格界限，是單邊規(guī)格，還是雙邊規(guī)格。目標(biāo)值是在中心，或則不在中心測(cè)量系統(tǒng)的分析要先做好。8STEP1決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP2決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明在收集Y特性時(shí)要注意層別和分組。各項(xiàng)的數(shù)據(jù)要按時(shí)間順序做好相應(yīng)的整理9STEP2決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP3決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明將數(shù)據(jù)輸入MINTAB中，或則在EXCEL中都可以。10STEP3決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP4決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明利用MINITAB>STAT>QUALITYTOOL>CAPABILITYANALYSIS(NORMAL)11STEP4決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP5決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明利用MINITAB的各項(xiàng)圖形來進(jìn)行結(jié)果說明12STEP5決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITA練習(xí)樣本X1X2X3X4X5199.7098.72100.24101.28101.20299.32100.97100.8799.2498.21399.8999.83101.4899.56100.90499.1599.7199.1799.3098.80599.66100.80101.06101.16100.45697.7498.8299.2498.6498.737101.18100.2499.6299.3399.918101.54100.96100.62100.67100.499101.49100.6799.36100.38102.101097.1698.2697.59100.0999.7813練習(xí)樣本X1X2X3X4X5199.7098.72100輸入數(shù)據(jù)14輸入數(shù)據(jù)14執(zhí)行能力分析15執(zhí)行能力分析15輸入選項(xiàng)16輸入選項(xiàng)16選擇標(biāo)準(zhǔn)差的估計(jì)方法17選擇標(biāo)準(zhǔn)差的估計(jì)方法17選項(xiàng)的輸入18選項(xiàng)的輸入18以Cpk,Ppk結(jié)果的輸出19以Cpk,Ppk結(jié)果的輸出19以Zbench方式輸出20以Zbench方式輸出20結(jié)果說明請(qǐng)學(xué)員按此圖形來說明該製程狀況21結(jié)果說明請(qǐng)學(xué)員按此圖形來說明該製程狀況21CapabilityAnalysis(Between/Within)22CapabilityAnalysis(Between/WCapabilityAnalysis(Weibull)此項(xiàng)的分析是用在當(dāng)制程不是呈現(xiàn)正態(tài)分佈時(shí)所使用。因?yàn)槿绻瞥滩皇钦龖B(tài)分佈硬用正態(tài)分佈來分析時(shí)，容易產(chǎn)生誤差，所以此時(shí)可以使用韋氏分佈來進(jìn)行分析，會(huì)更貼近真實(shí)現(xiàn)像。23CapabilityAnalysis(Weibull)此練習(xí)請(qǐng)使用同前之?dāng)?shù)據(jù)來進(jìn)行分析。上規(guī)格：103下規(guī)格：97規(guī)格中心：10024練習(xí)請(qǐng)使用同前之?dāng)?shù)據(jù)來進(jìn)行分析。24選韋氏分佈25選韋氏分佈25輸入相關(guān)參數(shù)26輸入相關(guān)參數(shù)26填入選項(xiàng)要求27填入選項(xiàng)要求27結(jié)果圖形28結(jié)果圖形28比較二者有何差異此二項(xiàng)誰更適合來解釋制程狀況。如果你是制程工程師你應(yīng)如何抉擇29比較二者有何差異此二項(xiàng)誰更適合來解釋制程狀況。29正態(tài)分佈適用性的判定可以使用Stat>basicstatistic>normalitytest但數(shù)據(jù)要放到同一個(gè)column中，所以必須針對(duì)前面的數(shù)據(jù)進(jìn)行一下處理30正態(tài)分佈適用性的判定可以使用30數(shù)據(jù)調(diào)整31數(shù)據(jù)調(diào)整31選擇執(zhí)行項(xiàng)目32選擇執(zhí)行項(xiàng)目32填寫選項(xiàng)33填寫選項(xiàng)33結(jié)果輸出34結(jié)果輸出34結(jié)果輸出(加標(biāo)0.5概率)35結(jié)果輸出(加標(biāo)0.5概率)35計(jì)量型製程能力分析總結(jié)一般的正態(tài)分佈使用CapabilityAnalysis(Normal)如果是正態(tài)分佈且其組內(nèi)和組間差異較大時(shí)可用CapabilityAnalysis(Between/Within)當(dāng)非正態(tài)分佈時(shí)則可以使用CapabilityAnalysis(Weibull)36計(jì)量型製程能力分析總結(jié)一般的正態(tài)分佈使用36CapabilitySixpack(Normal)複合了以下的六個(gè)圖形XbarR原始數(shù)據(jù)分佈直方圖正態(tài)分佈檢定CPK,PPK37CapabilitySixpack(Normal)複合了練習(xí)請(qǐng)以前面的數(shù)據(jù)來進(jìn)行相應(yīng)的CapabilitySixpack(Normal)練習(xí)38練習(xí)請(qǐng)以前面的數(shù)據(jù)來進(jìn)行相應(yīng)的CapabilitySixp選capabilitysixpack(normal)39選capabilitysixpack(normal)3輸入各項(xiàng)參數(shù)40輸入各項(xiàng)參數(shù)40選定判異準(zhǔn)則41選定判異準(zhǔn)則41選擇標(biāo)準(zhǔn)差估計(jì)方法42選擇標(biāo)準(zhǔn)差估計(jì)方法42考慮可選擇項(xiàng)43考慮可選擇項(xiàng)43結(jié)果輸出44結(jié)果輸出44CapabilitySixpack(Between/Within)複合了以下的六個(gè)圖形XbarR原始數(shù)據(jù)分佈直方圖正態(tài)分佈檢定CPK,PPK45CapabilitySixpack(Between/Wi同前練習(xí)及結(jié)果46同前練習(xí)及結(jié)果46CapabilitySixpack(Weibull)複合了以下的六個(gè)圖形XbarR原始數(shù)據(jù)分佈直方圖正態(tài)分佈檢定CPK,PPK47CapabilitySixpack(Weibull)複合結(jié)果輸出48結(jié)果輸出48二項(xiàng)分佈制程能力分析二項(xiàng)分佈只適合用在好，不好過，不過好，壞不可以用在0,1,2,3等二項(xiàng)以的選擇，此種狀況必須使用卜氏分佈。49二項(xiàng)分佈制程能力分析二項(xiàng)分佈只適合用在49示例數(shù)據(jù)在excel檔案中50示例數(shù)據(jù)在excel檔案中50選二項(xiàng)分佈制程能力51選二項(xiàng)分佈制程能力51填好各項(xiàng)的參數(shù)52填好各項(xiàng)的參數(shù)52選好控制圖的判異準(zhǔn)則53選好控制圖的判異準(zhǔn)則53填入選擇項(xiàng)54填入選擇項(xiàng)54結(jié)果及輸出55結(jié)果及輸出55結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋56結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋56卜氏分佈制程能力分析卜分佈只適合用在計(jì)數(shù)型，有二個(gè)以上的選擇時(shí)例如可以用在外觀檢驗(yàn)，但非關(guān)鍵項(xiàng)部份0,1,2,3等二項(xiàng)以的選擇，此種狀況必須使用卜氏分佈。57卜氏分佈制程能力分析卜分佈只適合用在57示例數(shù)據(jù)在excel檔案中58示例數(shù)據(jù)在excel檔案中58選卜氏分佈制程能力59選卜氏分佈制程能力59填好各項(xiàng)的參數(shù)60填好各項(xiàng)的參數(shù)60選好控制圖的判異準(zhǔn)則61選好控制圖的判異準(zhǔn)則61填入選擇項(xiàng)62填入選擇項(xiàng)62結(jié)果及輸出63結(jié)果及輸出63結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋64結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋64ExampleofCapabilityAnalysisforMultipleVariables(Nonnormal)

1

OpentheworksheetMNCAPA.MTW.2

ChooseStat>QualityTools>CapabilityAnalysis>MultipleVariable(Nonnormal).3

InVariables,enterWeight.4

CheckBYvariablesandenterMachine.4

InFitdatawith,chooseDistributionandthenselectLargestextremevalue.5

InLowerspec,enter27.InUpperspec,enter35.6

ClickOK.

65ExampleofCapabilityAnalysisTheprobabilityplotconfirmsthatthedatafollowslargestextremevaluedistribution.Formachine1,AD=0.335andP>0.25.Formachine2,AD=0.341andP>0.25.Thecapabilitystatisticsarebasedonthe0.5,99.87and0.13percentilesdenotedasX0.5,X0.9987,andX0.0013.Thepercentiles

arecalculatedusingtheparameterestimatesforthelargestextremevaluedistribution.Ppisdefinedastheratioofthespecificationrange(USL-LSL)tothepotentialprocessrange(X0.9987-X0.0013).Ppformachine1andmachinetwoare0.84and0.90respectively,indicatingthattheprobabilitythattheprocessproducesconformingfrozenfoodpacketsisslightlylessthan0.9974.66TheprobabilityplotconfirmsPPListheratioofX0.5-LSLtoX0.5-X0.0013.PPUistheratioofUSL-X0.5toX0.9987-X0.5.Formachine1,PPL=1.33andPPU=0.66,indicatingthatmorethan0.13percentpftheprocessoutputismorethantheupperspecificationlimit.Thisalsoindicatesthattheprocesshasmedianclosetothelowerspecificationlimit.Thisisalsoevidentinthehistogram.Machine2showsimilarresults.PpkistheminimumofPPUandPPL.Forbothmachines,highvalueofPpandlowvalueofPpkindicatethattheprocessmedianisoffthespecificationmidpoint.Thisalsoindicatesthatmorethan0.13percentoftheprocessoutputisoutsideatleastoneofthespecificationlimits.67PPListheratioofX0.5-LSLThePPM<LSL(1.03209)indicatesthatformachine1,1outof1millionisexpectedtofallbelowthelowerspecificationlimitof27oz.ThePPM>USL(10904)indicatesthatformachine1,10904outof1millionareexpectedtoexceedtheupperspecificationlimitof35oz.Machinetwoshowsimilarresults.Industryguidelinesdeterminewhethertheprocessiscapable.Agenerallyacceptedminimumvaluefortheindicesis1.33.Forbothmachinesthecapabilityindicesarelowerthan1.33.Theprocesstendstoputmorefoodinapackagethantheupperlimit.Themanufacturerneedstotakeimmediatestepstoimprovetheprocess.

68ThePPM<LSL(1.03209)indicaCCpkisameasureofpotentialcapability.ItisidenticaltotheCpkindexexceptthat,insteadofbeingcenteredattheprocessmeanallthetime,itiscenteredatthetargetwhengivenorthemidpointofthespecificationlimitswhenthespecificationlimitsaregiven.CCpkispreciselyCpkwhenoneofthespecificationlimitsandthetargetisnotgiven.

69CCpkisameasureofpotentialMINITAB之製程能力分析易騰涂順章70MINITAB之製程能力分析易騰涂順章1製程能力之分類計(jì)量型(基於正態(tài)分佈)計(jì)數(shù)型(基於二項(xiàng)分佈)計(jì)數(shù)型(基於卜氏項(xiàng)分佈)71製程能力之分類計(jì)量型(基於正態(tài)分佈)計(jì)數(shù)型(基於二項(xiàng)分佈)計(jì)MINITAB能力分析的選項(xiàng)(計(jì)量型)CapabilityAnalysis(Normal)CapabilityAnalysis(Between/Within)CapabilityAnalysis(Weibull)CapabilitySixpack(Normal)CapabilitySixpack(Between/Within)CapabilitySixpack(Weibull)72MINITAB能力分析的選項(xiàng)(計(jì)量型)CapabilityCapabilityAnalysis(Normal)該命令會(huì)劃出帶理論正態(tài)曲線的直方圖，這可直觀評(píng)估數(shù)據(jù)的正態(tài)性。輸出報(bào)告中還包含過程能力統(tǒng)計(jì)表，包括子組內(nèi)和總體能力統(tǒng)計(jì)。73CapabilityAnalysis(Normal)該命CapabilityAnalysis(Between/Within)該命令會(huì)劃出帶理論正態(tài)曲線的直方圖，可以直觀評(píng)估數(shù)據(jù)的正態(tài)性。該命令適用於子組間存在較變差的場(chǎng)合。輸出報(bào)告中還包含過程能力統(tǒng)計(jì)表，包括子組間／子組內(nèi)和總體能力統(tǒng)計(jì)。74CapabilityAnalysis(Between/WCapabilityAnalysis(Weibull)該命會(huì)會(huì)劃出帶韋伯曲線的直方圖，這可直觀評(píng)估數(shù)據(jù)是否服從韋伯分布。輸出報(bào)告中還包含總體過程總能力統(tǒng)計(jì)75CapabilityAnalysis(Weibull)該製程能力分析做法決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明76製程能力分析做法決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)STEP1決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明Y特性一般是指客戶所關(guān)心所重視的特性。Y要先能量化，儘量以定量數(shù)據(jù)為主。Y要事先了解其規(guī)格界限，是單邊規(guī)格，還是雙邊規(guī)格。目標(biāo)值是在中心，或則不在中心測(cè)量系統(tǒng)的分析要先做好。77STEP1決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP2決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明在收集Y特性時(shí)要注意層別和分組。各項(xiàng)的數(shù)據(jù)要按時(shí)間順序做好相應(yīng)的整理78STEP2決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP3決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明將數(shù)據(jù)輸入MINTAB中，或則在EXCEL中都可以。79STEP3決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP4決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明利用MINITAB>STAT>QUALITYTOOL>CAPABILITYANALYSIS(NORMAL)80STEP4決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITASTEP5決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITAB數(shù)據(jù)表進(jìn)行分析結(jié)果說明利用MINITAB的各項(xiàng)圖形來進(jìn)行結(jié)果說明81STEP5決定Y特性決定Y特性收集Y特性數(shù)據(jù)輸入MINITA練習(xí)樣本X1X2X3X4X5199.7098.72100.24101.28101.20299.32100.97100.8799.2498.21399.8999.83101.4899.56100.90499.1599.7199.1799.3098.80599.66100.80101.06101.16100.45697.7498.8299.2498.6498.737101.18100.2499.6299.3399.918101.54100.96100.62100.67100.499101.49100.6799.36100.38102.101097.1698.2697.59100.0999.7882練習(xí)樣本X1X2X3X4X5199.7098.72100輸入數(shù)據(jù)83輸入數(shù)據(jù)14執(zhí)行能力分析84執(zhí)行能力分析15輸入選項(xiàng)85輸入選項(xiàng)16選擇標(biāo)準(zhǔn)差的估計(jì)方法86選擇標(biāo)準(zhǔn)差的估計(jì)方法17選項(xiàng)的輸入87選項(xiàng)的輸入18以Cpk,Ppk結(jié)果的輸出88以Cpk,Ppk結(jié)果的輸出19以Zbench方式輸出89以Zbench方式輸出20結(jié)果說明請(qǐng)學(xué)員按此圖形來說明該製程狀況90結(jié)果說明請(qǐng)學(xué)員按此圖形來說明該製程狀況21CapabilityAnalysis(Between/Within)91CapabilityAnalysis(Between/WCapabilityAnalysis(Weibull)此項(xiàng)的分析是用在當(dāng)制程不是呈現(xiàn)正態(tài)分佈時(shí)所使用。因?yàn)槿绻瞥滩皇钦龖B(tài)分佈硬用正態(tài)分佈來分析時(shí)，容易產(chǎn)生誤差，所以此時(shí)可以使用韋氏分佈來進(jìn)行分析，會(huì)更貼近真實(shí)現(xiàn)像。92CapabilityAnalysis(Weibull)此練習(xí)請(qǐng)使用同前之?dāng)?shù)據(jù)來進(jìn)行分析。上規(guī)格：103下規(guī)格：97規(guī)格中心：10093練習(xí)請(qǐng)使用同前之?dāng)?shù)據(jù)來進(jìn)行分析。24選韋氏分佈94選韋氏分佈25輸入相關(guān)參數(shù)95輸入相關(guān)參數(shù)26填入選項(xiàng)要求96填入選項(xiàng)要求27結(jié)果圖形97結(jié)果圖形28比較二者有何差異此二項(xiàng)誰更適合來解釋制程狀況。如果你是制程工程師你應(yīng)如何抉擇98比較二者有何差異此二項(xiàng)誰更適合來解釋制程狀況。29正態(tài)分佈適用性的判定可以使用Stat>basicstatistic>normalitytest但數(shù)據(jù)要放到同一個(gè)column中，所以必須針對(duì)前面的數(shù)據(jù)進(jìn)行一下處理99正態(tài)分佈適用性的判定可以使用30數(shù)據(jù)調(diào)整100數(shù)據(jù)調(diào)整31選擇執(zhí)行項(xiàng)目101選擇執(zhí)行項(xiàng)目32填寫選項(xiàng)102填寫選項(xiàng)33結(jié)果輸出103結(jié)果輸出34結(jié)果輸出(加標(biāo)0.5概率)104結(jié)果輸出(加標(biāo)0.5概率)35計(jì)量型製程能力分析總結(jié)一般的正態(tài)分佈使用CapabilityAnalysis(Normal)如果是正態(tài)分佈且其組內(nèi)和組間差異較大時(shí)可用CapabilityAnalysis(Between/Within)當(dāng)非正態(tài)分佈時(shí)則可以使用CapabilityAnalysis(Weibull)105計(jì)量型製程能力分析總結(jié)一般的正態(tài)分佈使用36CapabilitySixpack(Normal)複合了以下的六個(gè)圖形XbarR原始數(shù)據(jù)分佈直方圖正態(tài)分佈檢定CPK,PPK106CapabilitySixpack(Normal)複合了練習(xí)請(qǐng)以前面的數(shù)據(jù)來進(jìn)行相應(yīng)的CapabilitySixpack(Normal)練習(xí)107練習(xí)請(qǐng)以前面的數(shù)據(jù)來進(jìn)行相應(yīng)的CapabilitySixp選capabilitysixpack(normal)108選capabilitysixpack(normal)3輸入各項(xiàng)參數(shù)109輸入各項(xiàng)參數(shù)40選定判異準(zhǔn)則110選定判異準(zhǔn)則41選擇標(biāo)準(zhǔn)差估計(jì)方法111選擇標(biāo)準(zhǔn)差估計(jì)方法42考慮可選擇項(xiàng)112考慮可選擇項(xiàng)43結(jié)果輸出113結(jié)果輸出44CapabilitySixpack(Between/Within)複合了以下的六個(gè)圖形XbarR原始數(shù)據(jù)分佈直方圖正態(tài)分佈檢定CPK,PPK114CapabilitySixpack(Between/Wi同前練習(xí)及結(jié)果115同前練習(xí)及結(jié)果46CapabilitySixpack(Weibull)複合了以下的六個(gè)圖形XbarR原始數(shù)據(jù)分佈直方圖正態(tài)分佈檢定CPK,PPK116CapabilitySixpack(Weibull)複合結(jié)果輸出117結(jié)果輸出48二項(xiàng)分佈制程能力分析二項(xiàng)分佈只適合用在好，不好過，不過好，壞不可以用在0,1,2,3等二項(xiàng)以的選擇，此種狀況必須使用卜氏分佈。118二項(xiàng)分佈制程能力分析二項(xiàng)分佈只適合用在49示例數(shù)據(jù)在excel檔案中119示例數(shù)據(jù)在excel檔案中50選二項(xiàng)分佈制程能力120選二項(xiàng)分佈制程能力51填好各項(xiàng)的參數(shù)121填好各項(xiàng)的參數(shù)52選好控制圖的判異準(zhǔn)則122選好控制圖的判異準(zhǔn)則53填入選擇項(xiàng)123填入選擇項(xiàng)54結(jié)果及輸出124結(jié)果及輸出55結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋125結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋56卜氏分佈制程能力分析卜分佈只適合用在計(jì)數(shù)型，有二個(gè)以上的選擇時(shí)例如可以用在外觀檢驗(yàn)，但非關(guān)鍵項(xiàng)部份0,1,2,3等二項(xiàng)以的選擇，此種狀況必須使用卜氏分佈。126卜氏分佈制程能力分析卜分佈只適合用在57示例數(shù)據(jù)在excel檔案中127示例數(shù)據(jù)在excel檔案中58選卜氏分佈制程能力128選卜氏分佈制程能力59填好各項(xiàng)的參數(shù)129填好各項(xiàng)的參數(shù)60選好控制圖的判異準(zhǔn)則130選好控制圖的判異準(zhǔn)則61填入選擇項(xiàng)131填入選擇項(xiàng)62結(jié)果及輸出132結(jié)果及輸出63結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋133結(jié)果解釋請(qǐng)針對(duì)前圖進(jìn)行相應(yīng)的各項(xiàng)解釋64ExampleofCapabilityAnalysisforMultipleVariables(Nonnormal)

1

OpentheworksheetMNCAPA.MTW.2

ChooseStat>QualityTools>CapabilityAnalysis>MultipleVariable(Nonnormal).3

InVariables,enterWeight.4

CheckBYvariablesandenterMachine.4

InFitdatawith,chooseDistributionandthenselectLargestextremevalue.5

InLowerspec,enter27.InUpperspec,enter35.6

ClickOK.

134ExampleofCapabilityAnalysisTheprobabilityplotconfirmsthatthedatafollowslargestextremevaluedistribution.Formachine1,AD=0.335andP>0.25.Formachine2,AD=0.341andP>0.25.Thecapabilitystatisticsarebasedonthe0.5,99.87and0.13percentilesdenotedasX0.5,X0.9987,andX0.0013.Thepercentiles

arecalculatedusingtheparameterestimatesforthelargestextremevaluedistribution.Ppisdefinedastheratioofthespecificationrange(USL-LSL)tothepotentialprocessrange(X0.9987-X0.0013).Ppformachine1andmachinetwoare0.84and0.90respectively,indicatingthattheprobabilitythattheprocessproducesconformingfrozenfoodpacketsisslightlylessthan0.9974.135TheprobabilityplotconfirmsPPListheratioofX0.5-LSLtoX0.5-X0.0013.PPUistheratioofUSL-X0.5toX0.9987-X0.5.Formachine1,PPL=1.33andPPU=0.66,indicatingthatmorethan0.13percentpftheprocessoutputismorethantheupperspecificationlimit.Thisalsoindicatesthattheprocesshasmedianclosetothelowerspecificationlimit.Thisisalsoevidentinthehistogr