1、統(tǒng)計(jì)過(guò)程控制工作坊 SPC Work Shop精益管理顧問(wèn)公司 -企業(yè)管理的好幫手 1基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept2統(tǒng)計(jì)學(xué)(Statistics)收集、整理、展示、分析、解析統(tǒng)統(tǒng)計(jì)資料Collect,collate,display,analysis,explainofdata由樣本推論母體體/群體From sampletoproject thepopulation能在不確確定情況況下作決決策Decisionwith uncertainty是一門(mén)科科學(xué)方法法、決策工具Scientific anddecisiontool基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept抽樣推論3統(tǒng)計(jì)量S

2、tatisticalValueR全距(range)算術(shù)平均均數(shù)(arithmetic mean)Md中位數(shù)(median)Mo眾數(shù)(mode)2方差/變異(variance)標(biāo)準(zhǔn)差(standarddeviation)基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept4R全距(range)R=Xmax-Xmin算術(shù)平均均數(shù)(arithmetic mean)基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept5Md中位數(shù)(median)順序數(shù)列列中的中中心項(xiàng)的的數(shù)值Themedial numberintheseriesMo眾數(shù)(mode)資料中出出現(xiàn)最多多的數(shù)值值Thenumber withhighestfr

3、equency基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept62方差/變異(variance)基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept7標(biāo)準(zhǔn)差(standarddeviation)基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept8例：Case1,1,2,3,4,6,11R=10=4Md=3Mo=1n2=10.86n-12=12.67n=3.29n-1=3.56基本統(tǒng)計(jì)計(jì)概念BasicSPCConcept9數(shù)列Series： 12,11,12, 13,18,30, 24,9請(qǐng)計(jì)算下下列統(tǒng)計(jì)計(jì)量Calculatethefollowing:RMdMon2=n-12=n=n-1=練習(xí)Exercise10

4、數(shù)據(jù)的收收集與整理Data Collectionand Collation分析測(cè)試行動(dòng)ActionTestAnalysis數(shù)據(jù)Data結(jié)Conclusion樣本Sample群體PopulationSampling抽 樣11數(shù)據(jù)收集Data Collection12數(shù)據(jù)整理理Data Collate-次數(shù)分配配FrequencyDistribution13直方圖(Histogram)14分布曲線DistributionCurve15正態(tài)(常態(tài))分布的曲曲線圖NormalDistribution Curve16正態(tài)分布布的性質(zhì)NatureofNormal Distribution17數(shù)據(jù)Data直

5、方圖Histogram分布曲線Distribution18直方圖Histogram分布Distribution正常型Normal19直方圖Histogram分布Distribution偏向型Bias20直方圖Histogram分布Distribution孤島型Island21過(guò)程變差Process Variation22輸入(材料)Input(Material)人Man方法Method過(guò)程(生產(chǎn)/裝配)Process輸出(產(chǎn)品)Output(Product)反饋(測(cè)量/檢驗(yàn))Feedback(Test/Insp.)測(cè)量系統(tǒng)統(tǒng)MSA)材料Material過(guò)程變差Process Variation

6、機(jī)器Machine23過(guò)程變差Process Variation輸入材料MaterialInput不同批次次之間的的差異LottoLot Variation批次內(nèi)的的差異WithinLot Variation隨時(shí)間產(chǎn)產(chǎn)生的差差異VariationinTimeSeries隨環(huán)境而而產(chǎn)生的的差異Variationdue to Environment24過(guò)程變差Process Variation生產(chǎn)/裝配Production/Assembly設(shè)備及工工裝夾具具的差異Equipmentand ToolingVariation隨時(shí)間而而產(chǎn)生的的磨損，漂移等AgingandShiftbytime操作工之之

7、間的差差異(如手工操操作的過(guò)過(guò)程)VariationbetweenOperators(ManualWork)設(shè)置的差差異DifferentSetup環(huán)境的差差異DifferentEnvironments25過(guò)程變差Process Variation輸出產(chǎn)品OutputProduct輸出的產(chǎn)產(chǎn)品隨時(shí)時(shí)間而產(chǎn)產(chǎn)生的變變化Product Variationbytime series輸出的產(chǎn)產(chǎn)品隨環(huán)環(huán)境而產(chǎn)產(chǎn)生的變變化Product Variationduetoenvironment26過(guò)程變差Process Variation反饋-測(cè)量Feedback-Measurement由于測(cè)量量用于過(guò)過(guò)程中的

8、的所有組組成部分分，測(cè)量量的變差差會(huì)對(duì)過(guò)過(guò)程的各各個(gè)階段段產(chǎn)生影影響Measurementsare partofthe process, variationinthemeasurementcan affecttheprocessinvarious stages偏倚Bias穩(wěn)定性Stability重復(fù)性Repeatability再現(xiàn)性Reproducibility分辨率Resolution27持續(xù)改進(jìn)進(jìn)過(guò)程循循環(huán)的各各個(gè)階段段2.維護(hù)過(guò)程程監(jiān)控過(guò)程程性能查找偏差差的特殊殊原因並採(cǎi)取措措施3.改進(jìn)過(guò)程程.改變過(guò)程程從而更更好地理理解普通通原因變差差減少普通通原因變變差分析過(guò)程程本過(guò)程應(yīng)應(yīng)做些什什麼

9、?會(huì)出現(xiàn)什什麼錯(cuò)誤誤?本過(guò)程正正在做什什麼?達(dá)到統(tǒng)計(jì)計(jì)控制狀狀態(tài)?確定能力力計(jì)劃實(shí)施措施研究計(jì)劃措施實(shí)施研究計(jì)劃實(shí)施措施研究28過(guò)程變差Process Variation范圍范圍范圍范圍每件產(chǎn)品品的尺寸寸與別的的不同Each productdimension hasdifferencesSPEC29過(guò)程變差Process Variation范圍范圍范圍但它們形形成一個(gè)個(gè)模型，若穩(wěn)定定，可以以描述為為一個(gè)分布Thedataaregroupedtoform themodel.Ifstable,itcanbedescriedasthe distribution.30過(guò)程變差Process Varia

10、tion分布可以以通用過(guò)過(guò)以下因因素來(lái)加加以區(qū)分Distributioncan be descried withvariouselements或這些因因數(shù)的組組合Orthecombinationoftheseelements范圍范圍范圍位置Position分布寬度度置Spread形狀Form/Shape31過(guò)程變差Process Variation預(yù)測(cè)Forecast范圍 SPEC時(shí)間 Time如果只存存在變差差的普通通原因，隨著時(shí)間間的推移移，過(guò)程程的輸出出形成一個(gè)個(gè)穩(wěn)定的的分布并并可預(yù)測(cè)測(cè)Ifthereareonlythecommon causes,and moveintimeseries

11、, theprocess outputsare predictable withstable distribution目標(biāo)直線線TargetLine32過(guò)程變差Process Variation目標(biāo)直線TargetLine預(yù)測(cè)Forecast范圍Spec時(shí)間Time如果存在在變差的的特殊原因，隨隨著時(shí)間間的推移移，過(guò)程程的輸出出不穩(wěn)定Ifthespecialcauseisinthevariation,intime series,the processbecome unstable？33過(guò)程控制制和過(guò)程程能力Process Controland ProcessCapability過(guò)程控制制Con

12、trol Process受控InControl(消除了特特殊原因因)Eliminatethe SpecialCause時(shí)間 Time范圍 SPEC不受制OutOfControl(存在特殊殊原因) withSpecialCause34過(guò)程變差Process VariationDistributioncan be describedby:Location(typicalvalue)位置(典型值值)Spread分布寬度度Shape(patternofvariation)形狀(變差的的模式)LSLUSLTypical value典型值Spread分布35過(guò)程變差Process Variation過(guò)程的

13、變變差分兩兩個(gè)類型TwoCategories特殊原因因造成的的變差Special Causes普通原因因造成的的變差CommonCauses36過(guò)程變差Process Variation普通原因因變差CommonCauses in Variation影響過(guò)程程中每個(gè)個(gè)單位affecteachunit in theprocess是由所有有不可分分派的小小變差源源組成Formedbyvariousvariationsources;and但遵循一一個(gè)分布followaparticular distribution通常需要采取取系統(tǒng)措措施來(lái)減減小Usesystem correctiveactionst

14、oreduce在控制圖圖上表現(xiàn)現(xiàn)為隨機(jī)機(jī)性randompointinthe controlchartIfonly commoncause,the processoutput of aprocessisPredictable如只有普普遍原因因，輸出出是可預(yù)預(yù)計(jì)的過(guò)過(guò)程。37過(guò)程變差差Process Variation特殊原因因變差Special CausesinVariation間斷的，偶然的的，通常常是不可可預(yù)測(cè)的的和不穩(wěn)穩(wěn)定的變變差Discontinued, accidental, unpredictableand unstable非隨機(jī)的的圖案Nonrandom pattern是由可分分派

15、的變變差源造造成該變變差源可可以被糾糾正Canbecorrected,thecauseisfromtheparticularsource38過(guò)程變差Process Variation工業(yè)經(jīng)驗(yàn)驗(yàn)建議為為：Industrial Experience局部措施施Localaction通常用來(lái)來(lái)消除變變差的特殊原因因Usually required to eliminatespecial causesofvariation通常由與與過(guò)程直直接相關(guān)關(guān)的人員員實(shí)施Usually taken by peopleclosetotheprocess大約可糾糾正15%的過(guò)過(guò)程問(wèn)題題Only 15%ofprocess

16、 problemcan be corrected39過(guò)程變差Process Variation工業(yè)經(jīng)驗(yàn)驗(yàn)建議為為：Industrial Experience對(duì)系統(tǒng)采采取措施施Actions on thesystem通常用來(lái)來(lái)消除變變差的普通原因因Usually required to eliminate/ orreducecommon causesofvariation, andbetakenbymanagement大約可糾糾正85%的過(guò)過(guò)程問(wèn)題題About85%ofprocessproblem canbecorrected40過(guò)程控制制和過(guò)程程能力Process Controland Pro

17、cessCapability過(guò)程能力Process Capability規(guī)范上限受控且有有能力符符合規(guī)范范(普通原原因造成成特變差差)Control andcapable to fitthespec(variation duetoNormalCause)受控但沒(méi)沒(méi)有能力力符合規(guī)規(guī)范Control butwithout capabilitytofit thespec(普通原因因造成變變差太大大)Largevariationdue to NormalCauseSpec時(shí)間Time41過(guò)程控制制和過(guò)程程能力Process Controland ProcessCapability每個(gè)過(guò)程程可以分分類如

18、下下：Each processcan be categorized as:Control or outofcontrol ?受控或不不受控?Fitthe customer requirements?是否有滿滿足客戶戶要求?Meeting requirement滿足要求In Control受控Not In Control 不受控Acceptance符合Case 1 第1類Case 3 第3類Not Acceptance不符合Case 2 第2類Case 4 第4類42過(guò)程控制制和過(guò)程程能力第1類(符合要求求，受控控)Case1 (Accept, In Control)是理想狀狀況。為為持續(xù)改改進(jìn)

19、可能能需要進(jìn)進(jìn)一步減減小變差差I(lǐng)dealStatus, to reducevariationfor continuousimprovement.第2類(不符合要要求，受受控)Case2 (NotAccept,InControl)存在過(guò)大大的普通通原因變變差with large variationfrom normalcause短期內(nèi)，進(jìn)行100%檢測(cè)以保保護(hù)客戶戶不受影影響Inshortterm,gofor100%inspection to reducetheeffect forcustomer必須進(jìn)行行持續(xù)改改善以減少普通原因因的變差Must go forcontinuous improve

20、ment to eliminatetheeffect fromcommon cause43過(guò)程控制制和過(guò)程程能力Process Controland ProcessCapability第3類(符合要求求，不受受控)Case3.Accept butOUTofcontrol有相對(duì)較較小的普普通原因因及特殊殊原因變變差With lessnormal andspecial causes如果存在在的特殊殊原因已已經(jīng)明確確但消除除具影響響可能不不大經(jīng)濟(jì)濟(jì)，客戶戶可能接接受這種種過(guò)程狀狀況Ifthespecialcausesare known butthecostmaybehightoeliminateth

21、em,customermay acceptthis processstatus.44過(guò)程控制制和過(guò)程程能力第4類(不符合要要求，不不受控)Case 4(NotAccept,OUTofControl)存在過(guò)大大的普通通原因及及特殊原原因的變變差LargerNormal andspecial causesinthevariation需要進(jìn)行行100%檢測(cè)以保保護(hù)客戶戶利益必必須采采取緊急急措施使使過(guò)程穩(wěn)穩(wěn)定，并并減小變變差Must go forcontingency/containmentactionstostablethe processand reducevariations,100% ins

22、pectiontoprotectcustomer45過(guò)程控制制和過(guò)程程能力判斷一個(gè)個(gè)過(guò)程是是否滿足足規(guī)格要要求：Judgetheprocesscanfit thespec requirementsCapability Index能力指數(shù)- Cpk(For stableProcess only只用於穩(wěn)穩(wěn)定過(guò)程)PerformanceIndex性能指數(shù) Ppk判斷一個(gè)個(gè) 過(guò)程程是受控還是不受控用：Judgetheprocessisincontrol or outofcontrol控制圖Control Charts46控制圖Control Charts為判斷制制程變差差的普通原因因或特殊殊原因的主要

23、統(tǒng)統(tǒng)計(jì)分析析工具TojudgetheprocessvariationwithCOMMONorSPECIALcauseasthemajorstatisticalanalysistool47基本概念BasicConcept控制圖中中包括三三條線3 lines in ControlChart控制上限限(UpperControl Limit;UCL)中心線(Center Line;CL)控制下限限(LowerControlLimit;LCL)UCLCLLCL48控制圖的的種類Type of ControlCharts計(jì)量型數(shù)數(shù)據(jù)的控控制圖VariableControl ChartsXbar-RXbar

24、-SX-MR計(jì)數(shù)型數(shù)數(shù)據(jù)的控控制圖AttributeControlChartspnpcu49控制圖的的選用程程序注 :本本圖假假設(shè)測(cè)量量系統(tǒng)已已經(jīng)過(guò)評(píng)價(jià)並並且是適適用的50控制圖的的使用策策劃Planningforthe ControlCharts作控制圖圖需要按按以下步步驟：Followthe necessarysteps:計(jì)劃Planning資源Resources評(píng)估和改改進(jìn)Evaluation andImprovement51控制圖的的使用策策劃Preparationforthe useofControlCharts建立適于于采取措措施的環(huán)環(huán)境-公司文化化Setthe right env

25、ironment to takeactionsCompany Culture確定過(guò)程程Definethe process確定待管管理的特特性(如:特特殊特性性)，考慮Determinecharacteristics (e.g. SpecialCharacteristics)that needtobemanagedShallConsider:客戶的要要求CustomerRequirements當(dāng)前及潛潛在的問(wèn)問(wèn)題區(qū)Existingandpotential problemarea特性間的的相互關(guān)關(guān)系Interrelation betweencharacteristics確定測(cè)量量系統(tǒng)Definet

26、he Measurement SystemMSA使不必要要的變差差最小化化Reducethe variationsinthoseavoidable area52控制圖的的使用策策劃Preparationforthe useofControlCharts過(guò)程 Process輸出 Output輸入 Input使用什么么方式？(材料料/設(shè)備備)What?(Equipment/RawMaterial)誰(shuí)進(jìn)行?(能力/技能/培訓(xùn))Who?(Competence,ability,training)如何做？(方法法/程序序/技術(shù)術(shù))How?(Method/Procedure/Technique)使用的關(guān)關(guān)鍵

27、準(zhǔn)則則是什么么？(測(cè)量/評(píng)估)Keymeasurement?(Measuring/Evaluation)53Xbar-RChart圖Average andRangeChart平均數(shù)-全距控制圖工業(yè)界最最常使用用的計(jì)量量值控制制圖Themostcommonchartfor theindustry54建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts收集數(shù)據(jù)據(jù)Collect Data選擇子組組大小，頻率和和數(shù)據(jù)Selectthe subgroup size, frequencyandnumbersofsubgroup子組大小小Subgroupsize :一個(gè)子組組

28、內(nèi)該單單元之間間出現(xiàn)變變差機(jī)會(huì)會(huì)小(如如45件連續(xù)續(xù)于單一一過(guò)程的的產(chǎn)品)variationamongunitswithin thesubgrouparesmall(e.g.4to5 consecutivelyproducedpieces by singleproduction stream)每個(gè)子組組內(nèi)的變變差是普普通原因因造成Variationwithin eachsubgroupreflectsCOMMON cause代表短時(shí)時(shí)間內(nèi)零零件間變變產(chǎn)Representthe piece-to-piece variationwithinshortperiod of time不正常變變差=過(guò)程發(fā)

29、發(fā)生變化化unusual variationbetween subgroups= changesinprocess55建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts選擇子組組大小，頻率和和數(shù)據(jù)Selectthe subgroup size, frequencyandnumbersofsubgroup子組頻率率Subgroupfrequency:檢查過(guò)程程經(jīng)過(guò)一一段時(shí)間間的變化化，如: 換班班、溫升升趨勢(shì)、材料變變更Todetectthe changeofprocess overtime,e.g.changes duetowork shift differe

30、nce, warm-uptrend, material lots56子組數(shù)大大小NumberofSubgroups :確保變差差的主要要原因有有機(jī)會(huì)出出現(xiàn)(25子組組包含100個(gè)個(gè)單值)Majorsources of variationhave hadanopportunitytoappear(25 or moresubgroups containing100 or morereadings)可利用現(xiàn)現(xiàn)有數(shù)據(jù)據(jù)加速研研究進(jìn)度度Canuse existing datatospeedupthe process建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts57建

31、立控制圖及記錄原原始數(shù)據(jù)據(jù)Setupcontrolchartandraw datagathering建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts58計(jì)算每個(gè)個(gè)子組的的X平均值和和極差RCalculatethe AverageRangeandProcess average建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts59UCLCLLCL子組數(shù)足夠時(shí)(25子子組包含含100個(gè)單值值),計(jì)算及繪繪成控制制線 試驗(yàn)驗(yàn)控制限限When numberofsubgroupisadequate(25ormoresubgroupsco

32、ntaining100ormorereadings),calculateand drawthe controllimittrialcontrol limits建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts60建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts過(guò)程控制制解釋Explain theProcess Control分析極差差圖(R圖)上的數(shù)據(jù)據(jù)點(diǎn)Analyse theRchartfirst識(shí)別并標(biāo)標(biāo)注特殊殊原因(R圖)Identifyandmarkthespecialcauses糾正及預(yù)預(yù)防再發(fā)發(fā)生Correct

33、andprevent it formrecurring消除受已識(shí)別的特殊原因因影響的子子組Exclude allsubgroupsaffectedbytheidentified specialcause重新計(jì)算算控制極極限(R圖)RecalculatetheRcontrol limits61當(dāng)極差受控過(guò)程分佈佈是穩(wěn)定定的,分析平均均值圖(Xbar圖)When therangesare in statistical control, processspread isstable, then看過(guò)程的的位置是是否改變變x圖Checkiftheprocesslocationischangingover

34、timexchart為繼續(xù)控控制延長(zhǎng)長(zhǎng)控制限Extendthe ControlLimits forProcess Control建立Xbar-R控制圖的的步驟Setupthe steps forX-RCharts62控制圖-過(guò)程的聲聲音Control Charts-the voice of theprocess控制圖可可以給我我們提供供出出現(xiàn)現(xiàn)了哪種種類型的的變差的的線索，供我們們采取相相應(yīng)的措施Thecontrolchartcanprovideusthehintsofwhattypesofvariations,forthe right correctiveactions63控制圖上上的信號(hào)號(hào)

35、解釋Explain TheSignals on theControl Charts有很多信信號(hào)規(guī)則則適用于于所有的的控制圖圖(Xbar圖和R圖)，主要最常常見(jiàn)的規(guī)規(guī)則有有以下3種：Many types of signalsand rules in theXRcharts, mainly3 types:規(guī)則1：超出控制制限的點(diǎn)Rule #1:PointsExceed thecontrol limitsUCLLCL64控制圖上上的信號(hào)號(hào)解釋Explain TheSignals on theControl Charts規(guī)則2：連續(xù)7點(diǎn)在中心心線一側(cè)Rules#2:continuous7 points

36、inonesideofthecentre linerunUCLLCL65控制圖上上的信號(hào)號(hào)解釋Explain TheSignals on theControl Charts規(guī)則3：連續(xù)7點(diǎn)升或下降Rule #3:continuous 7points in descendingorraisingtrendUCLLCL66過(guò)制圖上上的信號(hào)號(hào)解釋Explain TheSignals on theControl Charts多于2/3的點(diǎn)落在圖中1/3以外。More than 2/3 of points outside the 1/3 zone 67過(guò)制圖上上的信號(hào)號(hào)解釋Explain TheSign

37、als on theControl Charts呈現(xiàn)有規(guī)律變化。 Regular Cyclical Trend68改變控制制時(shí)限ChangeInControlLimit將這些的的控制限限畫(huà)到控控制圖上上，并作作為不斷斷進(jìn)行的的過(guò)程制的基礎(chǔ)礎(chǔ)。Plot these newcontrol limitsonthechartasthe basis forongoing processcontrol只要過(guò)程程的均值值和極差差保持受受控可可將將控制限限延長(zhǎng)以以于以后后的時(shí)期期Aslong as theprocess remainsincontrolforbothandrangeongoing limits

38、canbeextendedfor additionalperiods.69圖形 Patterns極差圖 Range Chart均值圖 Average Chart超出控制限的點(diǎn) Points beyond the control limits控制限繪錯(cuò) 分佈寬度增加(零件間變化) 測(cè)量系統(tǒng)未有足夠分辨或已變化Mis-plotted of control limits Increase or decrease of spread of distribution (piece-to-piece variation) Change in Measurement system or lack of di

39、scrimination 控制限繪錯(cuò) 過(guò)程已變化 測(cè)量系統(tǒng)己變化Mis-plotted of control limits Process has shifted Change in Measurement system 70圖形 Patterns極差圖 Range Chart均值圖 Average ChartRun 輸出值的分佈寬度增加： 無(wú)規(guī)律的成因，如設(shè)備異常或夾具鬆動(dòng) 測(cè)量系統(tǒng)改變 Greater spread of output value due to: irregular cause such as equipment malfunction or loose fixturing

40、 shift in process elements such as new , less uniform raw material lots Change in Measurement system 過(guò)程均值已變化 測(cè)量系統(tǒng)逐漸改變Process average has changed Change in Measurement system such as drift 71 圖形 極差圖 均值圖 明顯非隨機(jī)圖形 顯著多於2/3以上的點(diǎn)落在中線外： 控制限繪錯(cuò) 過(guò)程或抽樣方法會(huì)被分層：子組系統(tǒng)化包括兩個(gè)或以上的過(guò)程流過(guò)程均值變化 數(shù)據(jù)被編輯 顯著少於2/3以上的點(diǎn)落在中線外： 過(guò)程或抽樣方法

41、形成連續(xù)的分組中包括多個(gè)具不同變化的過(guò)程流 顯著多於2/3以上的點(diǎn)落在中線外： 控制限繪錯(cuò) 過(guò)程或抽樣方法會(huì)被分層：子組系統(tǒng)化包括兩個(gè)或以上的過(guò)程流過(guò)程均值變化 數(shù)據(jù)被編輯 顯著少於2/3以上的點(diǎn)落在中線外： 過(guò)程或抽樣方法形成連續(xù)的分組中包括多個(gè)具不同變化的過(guò)程流 如果存在幾個(gè)過(guò)程流，應(yīng)分別識(shí)別和追蹤 72改變控制制時(shí)限ChangeInControlLimit估計(jì)過(guò)程程的標(biāo)準(zhǔn)準(zhǔn)偏差用用現(xiàn)在的的子組容容量計(jì)算算：Estimatetheprocessstandarddeviation. Using theexistingsubgroupsize calculate:=R/d2式中R為子組極極差的

42、均均值(在在極差受受控時(shí)期期)，d2隨樣本容容量變化化的常數(shù)數(shù)：WhereR is theaverage of thesubgroupranges(forperiods withthe rangesincontrol)andd2isa constant varyingbysample size:n2345678910d21.131.692.062.332.532.702.852.973.0873改變控制制時(shí)限ChangeInControlLimit按照新的的子組容容量查表表得到系系數(shù)d2、D3、D4和A2，計(jì)算新的的極差和和制限：Usingthetabled factorsfor d2,D3,

43、D4and A2 based on thenewsubgroupsize,calculate thenewrangeand controllimits:=Rnewd2UCLR=D4RnewLCLR=D3RnewUCLX=X +A2RnewUCLX=X -A2Rnew74改變控制制時(shí)限ChangeInControlLimit將這些的的控制限限畫(huà)到控控制圖上上，并作作為不斷斷進(jìn)行的的過(guò)程控控制的基基礎(chǔ)。Plot thenewcontrollimitsonthe chart as thebasisforongoingprocess control.只要過(guò)程程的均值值和極差差保持受受控可可將控控制圖的

44、的目的是是合理、經(jīng)濟(jì)的的控制狀狀態(tài)Aslong as theprocess remainsincontrolforbothaveragesandrangesOngoing limitscanbeextendedfor additionalperiods75有關(guān)控控制的的最后概概念用用于進(jìn)一一步的考考慮在一個(gè)生生產(chǎn)過(guò)程程中永遠(yuǎn)遠(yuǎn)無(wú)法達(dá)達(dá)到一種種完美的的控制狀狀態(tài)。A perfectstateofcontrolisneverattainable in aproductionprocess.過(guò)程控制制圖的目目的是合合理、經(jīng)經(jīng)濟(jì)的控控制狀態(tài)態(tài)。Thegoaloftheprocesscontrol ch

45、artsisreasonable andeconomicalstateofcontrol.如果一張張控制圖圖上從來(lái)來(lái)不出現(xiàn)現(xiàn)失控點(diǎn)點(diǎn)，我們們將嚴(yán)肅肅地查問(wèn)問(wèn)該操作作是否應(yīng)應(yīng)畫(huà)圖圖。Ifa chart never wentout of controlwewouldseriously questionwhether thatoperation shouldbecharted.對(duì)于車間間目的來(lái)來(lái)說(shuō)，一一個(gè)受控控的過(guò)程程即是僅僅有很少少百分比比的點(diǎn)失失控并并且對(duì)失失控點(diǎn)采采取過(guò)適適當(dāng)?shù)拇氪胧orshoppurposesa controlledprocessisconsidered to be on

46、ewhereonly asmall percentageofthe pointsgooutofcontrolandwhereout-of-controlpointsarefollowedbyproperaction.76請(qǐng)完成Xbar-R控制圖的的制作練練習(xí)PleaseCompleteControl Chart Drawing練習(xí)Exercise77控制圖-過(guò)程的聲聲音控制圖可可以區(qū)分分出普通通原因變變差和特特殊原因因變差Control Chart candistinguishthenormal causesandthe specialcauses特殊原因因變差要要求立即即采取措施Specia

47、l Causesneedsimmediateactions減少普通通原因變變差需要要改變產(chǎn)產(chǎn)品或過(guò)過(guò)程的設(shè)設(shè)計(jì)ToReducethe normalcausesneedproduct or processredesign78控制圖-過(guò)程的聲聲音錯(cuò)誤的措施WrongActions試圖通過(guò)過(guò)持續(xù)調(diào)調(diào)整過(guò)程程參數(shù)來(lái)來(lái)固定住住普通原原因變差差，稱為為過(guò)度調(diào)調(diào)整，結(jié)結(jié)果會(huì)導(dǎo)導(dǎo)致更大大的過(guò)程程變差造造成客戶戶滿意度度下降Continuetoadjustthe processparameterstofixthe normalprocess variations- overadjustmentsonlycaus

48、elargervariationsandcustomerunsatisfied試圖通過(guò)過(guò)改變?cè)O(shè)設(shè)計(jì)來(lái)減減小特殊殊原因變變差可能能解決不不了問(wèn)題題，會(huì)造造成時(shí)間間和金錢錢的浪費(fèi)Trytochange designtoreducethe specialcauses in thevariations cannotsolveproblems,onlywastemoneyandtime79過(guò)度調(diào)整整是把每每一個(gè)偏偏離目標(biāo)標(biāo)的值當(dāng)當(dāng)作過(guò)程程中特殊殊原因處處理的作作法Over adjustmentisthe practice of treating eachdeviation fromthe targeta

49、sifitwere theresultofthe actionofa specialcauseofvariation in theprocess.Results withnoadjustment沒(méi)有經(jīng)過(guò)過(guò)調(diào)整的的結(jié)果NormalVariation正常變差差80Note IncreaseinVariation注意變差差增加Resultwithadjustment to compensatefor lastdeviation fromtarget對(duì)最終目目標(biāo)差補(bǔ)補(bǔ)償調(diào)整整後的結(jié)結(jié)果Results withadjustmenttocompensate forlast deviationfrom ta

50、rgetifdeviationwasgreater than1偏差大於於目標(biāo)1個(gè)單位位時(shí),進(jìn)進(jìn)行補(bǔ)補(bǔ)償調(diào)整整後的結(jié)結(jié)果Note IncreaseinVariation注意變差差增加81Xbar-S圖Chart平均數(shù)-標(biāo)準(zhǔn)差控制圖Average andStd. DevControl Chart與Xbar-R圖相同Same as X&RchartsS圖檢出能能力較R圖大，但但計(jì)算大大麻煩S canhave betterdetectionthanR,trouble in calculation一般樣本本大小n小于是9使用R圖，n大于9使用S圖SamplesizeN9,useScharts配合電腦使用

51、UsewithPC82Xbar-S圖Chart83X- MR圖Chart個(gè)別值-移動(dòng)全距距控制圖MovingRangeControlCharts數(shù)據(jù)不能能合理分分組時(shí)使用Data notsuitableforthe subgrouping在下列情情況可使使用：Inthefollowing situation:一次只能能收集到到一個(gè)數(shù)數(shù)據(jù)，生生產(chǎn)效率率及損耗耗率。Cancollecta singledata pertime,byconsideringtheefficiencyandscrap制程品質(zhì)質(zhì)極為均均勻，不不需多取取樣本，如液體體濃度。Theproductitselfisveryeven

52、lydistributed, no moresampling, likeConc.取得測(cè)定定值既費(fèi)費(fèi)時(shí)成本本又高，如復(fù)雜雜的化學(xué)學(xué)分析及及破壞性性試驗(yàn)。Costlyfor themeasurement,suchascomplicatedchemicalanalysis,and destructive test.84X- MR圖85計(jì)數(shù)型數(shù)數(shù)據(jù)控制制圖AttributeControlChartsP圖(PCharts)nP圖(nPCharts)C圖(CCharts)U圖(Ucharts)86P圖(PCharts)不合格品品率(不良率)控制圖Failure RateControlChart分析或控控

53、制制程程不合格品品率(不良率)Analyse or controlthe failurerate樣本大小小n可以不同同Samplesizen canvaries87P圖(PCharts)88P圖(PCharts)89nP圖(nPCharts)不合格品品數(shù)(不良數(shù))控制圖Nonconformingno( No of defective)ControlChart分析或控控制制程程不合格品品數(shù)(不良數(shù))Analyse andControl of Nonconforming no (Noofdefective)Control Chart樣本大小小n要相同SampleSizeN mustbethe sa

54、me90nP圖(nPCharts)91C圖C Charts缺點(diǎn)數(shù)控制圖DefectPointControlChart分析或控控制制程程缺點(diǎn)數(shù)數(shù)Analyse or ControlofDefect Point ControlChart樣本大小小n要相同SampleSizeN mustbethe same92C圖-在樣本大小小一致的的子組內(nèi)內(nèi)的不合格數(shù)93u圖U Charts單位缺點(diǎn)點(diǎn)數(shù)控制制圖Unit defectiveControl Charts分析或控控制制程程單位缺缺點(diǎn)數(shù)Analyse or ControlofUnitdefectiveControlCharts樣本大小小n可以不同Samp

55、leSizen canvaries94u圖U charts95過(guò)程能力Process Capability96過(guò)程控制制和過(guò)程程能力Process Control&Process Capability判斷一個(gè)個(gè) 過(guò)程程是受控還是不受控用：Tojudgetheprocessareincontroloroutofcontrol:控制圖Control Chart判斷一個(gè)個(gè)過(guò)程是是否滿足足規(guī)格要要求：Tojudgetheprocesscanfulfillthespecification:能力指數(shù)數(shù) Cpk( ProcessCapability)性能指數(shù)數(shù)- Ppk( Performance Capab

56、ility)97短期ShortTerm長(zhǎng)期Long Term過(guò)程能力Process Capability短期研究 長(zhǎng)期研究一月星期一星期二二月ShortTermLong Term98過(guò)程能力力的研究Process CapabilityStudy短期普通原因特殊原因因短期能力力研究-過(guò)程性能ShortTerm-PerformanceCapability699過(guò)程能力力的研究Process CapabilityStudy過(guò)程性能能(ProcessPerformance)過(guò)程總變變差特殊原因因變差+普通原因因變差TotalProcess VariationduetobothSpecial andCommonCause由估估計(jì)計(jì)出Estimateform Sigmas100過(guò)程能力力的研究Process CapabilityStudy輸入 (材料)過(guò)程 (生產(chǎn)/裝配)輸出 (產(chǎn)品)反饋 (測(cè)量/檢驗(yàn))長(zhǎng)期能力力研究-過(guò)程能力Long TermCapabilityStudy-ProcessCapabilityInputMaterialProcess (Production)Output(Product)FeedbackMeasure101長(zhǎng)期能力力