SPC统计过程控制培训(1)15674.pptx

《SPC统计过程控制培训(1)15674.pptx》由会员分享，可在线阅读，更多相关《SPC统计过程控制培训(1)15674.pptx（133页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、1 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlStatistical Process Control Training 统统 计计 过过 程程 控控 制制 培培 训训TV Rheinland Academy ChinaTV Rheinland Academy China德国莱茵技术监督协会中国学术部德国莱茵技术监督协会中国学术部德国莱茵技术

2、监督协会中国学术部德国莱茵技术监督协会中国学术部2 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlnProgram Manager of TProgram Manager of TV Rheinland Academy ChinaV Rheinland Academy China德国莱茵技术监督协会中国学术部课程经理德国莱茵技术监督协会中国学术

3、部课程经理nRegistered Quality Consultant,TRegistered Quality Consultant,TV RheinlandV Rheinland德国莱茵技术监督协会注册质量顾问德国莱茵技术监督协会注册质量顾问nRegistered Senior Quality Auditor,Quality Society of AustralasiaRegistered Senior Quality Auditor,Quality Society of Australasia澳大利亚新西兰质量协会注册高级质量审核员澳大利亚新西兰质量协会注册高级质量审核员nCertifie

4、d Auditor VDA6.1,VDA-QMC/Germany Certified Auditor VDA6.1,VDA-QMC/Germany 德国汽车工业联合会注册德国汽车工业联合会注册VDA6.1VDA6.1认证审核员认证审核员Qualification for Tutor Mr.Tony Hung讲师洪讲师洪 源源3 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical

5、Process Control持续改进及统计过程概述持续改进及统计过程概述一收集数据并用统计方法来解释它们并不是最终目标，最终目标是对过程不断加深理解。4 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control持续改进及统计过程概述持续改进及统计过程概述二研究变差和应用统计知识来改进性能的基本概念适用于任何领域，可以是在车间或办公室。例如：机器（性能特

6、性）、记帐（差错率）、浪费分析（废品率）等。5 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control持续改进及统计过程概述持续改进及统计过程概述三SPC代表统计过程控制。应用统计技术控制输出（例如零件）应仅仅是第一步，只有当产出输出的过程成为我们努力的重点，这些方法才能在改进质量，提高生产率，降低成本上发挥作用。6 Copy right reserv

7、ed by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control持续改进及统计过程概述持续改进及统计过程概述四现有已完成的例子仅是为说明如何运用SPC，要真正理解SPC需要进一步与过程实际联系，现有的过程信息不能代替实际工作经验。7 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal

8、Training Form Statistical Process ControlStatistical Process Control持续改进及统计过程概述持续改进及统计过程概述8 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control持续改进及统计过程概述持续改进及统计过程概述六测量系统对合适的数据分析来说很重要，而且在收集过程数据之前就应很好地了

9、解它们，如果这样的一个系统缺少统计控制或他们的变差占过程数据总变差中很大比例，就可能作出不合适的决定。9 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control 人人 料料 环环 法法设 备 工 装测量系统MSA 机机 器器 产品产品过程过程Cpk过程能力过程能力Cmk机器能力机器能力10 Copy right reserved by TV Rhein

10、land Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control影响测量结果的因素影响测量结果的因素11 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control测量系统分析测量系统

11、分析 MSA测量系统误差的类型测量系统误差的类型n偏倚偏倚 (Bias)12 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Controln稳定性稳定性 (Stability)测量系统分析测量系统分析 MSA测量系统误差的类型测量系统误差的类型13 Copy right reserved by TV Rheinland Academy Shanghai Ed

12、ited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Controln线性线性 (Linearity)测量系统分析测量系统分析 MSA测量系统误差的类型测量系统误差的类型14 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Controln

13、重复性重复性 (Repeatability)测量系统分析测量系统分析 MSA测量系统误差的类型测量系统误差的类型由同一操作者对同一部件用同一测量仪器的由同一操作者对同一部件用同一测量仪器的多次测量多次测量15 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Controln再现性再现性 (Reproducibility)测量系统分析测量系统分析 MSA测量系统

14、误差的类型测量系统误差的类型由不同操作者对同一部件用同一测量仪器的测量由不同操作者对同一部件用同一测量仪器的测量16 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Controln量具重复性和再现性量具重复性和再现性Gage R&R(repeatability and reproducibility)测量系统分析测量系统分析 MSA测量系统研究测量系统研究n

15、适用于所有列入控制计划的测量系统适用于所有列入控制计划的测量系统v计量型计量型 (Variable)v计数型计数型 (Attribute)17 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control18 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.In

16、ternal Training Form Statistical Process ControlStatistical Process Control19 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control过程控制系统20 Copy right reserved by TV Rheinland Academy Shanghai Edited by

17、Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control过程控制系统21 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control过程控制系统22 Copy right reserved by TV Rheinland Academy S

18、hanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control过程控制系统对过程采取措施对重要的性能（过程或输出）采取措施从而避免它们偏离目标值太远是很经济的。采取的措施包括改变操作或改变过程本身更基本的因素，并追踪结果。23 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Proc

19、ess ControlStatistical Process Control过程控制系统对输出采取措施如果输出不能满足顾客的要求，可能将所有产品进行分类报废不合格品或返工，因此仅对输出检测并纠正不符合规范的产品，而没有分析过程中的根本原因是不经济的。24 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control变差的普通原因及特殊原因变差的普通原因及特殊

20、原因 普通原因是指造成随着时间的推移具有稳定的且可重复的分布过程中的许多变差的原因，其表现为一个稳定系统的偶然原因。25 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control变差的普通原因及特殊原因变差的普通原因及特殊原因 特殊原因是指造成不是始终作用于过程变差的原因，即当它们出现是将造成（整个）过程的分布，将用不可预测的方式来影响过程的输出，过程的

21、输出将不稳定。26 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control27 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Pr

22、ocess ControlElectrical Electrical Component LengthComponent LengthUpperSpecificationLimitLower SpecificationLimitActual MicrometerActual Micrometer Measurements Measurements实际测量结果实际测量结果实际测量结果实际测量结果No Less Than 9.5mmNo More Than 10.5mm28 Copy right reserved by TV Rheinland Academy Shanghai Edited by

23、 Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlElectrical Electrical Component LengthComponent LengthQuite some Quite some VariationVariation-Ending up as-Ending up as ScrapScrap一定程度的偏差一定程度的偏差一定程度的偏差一定程度的偏差-导致废品导致废品导致废品导致废品 UpperSpecificationLimitLower Specif

24、icationLimitScrapScrapNo Less Than 9.5mmNo More Than 10.5mm29 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control30 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Tra

25、ining Form Statistical Process ControlStatistical Process Control31 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control32 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Intern

26、al Training Form Statistical Process ControlStatistical Process Control局部措施和对系统采取措施局部措施和对系统采取措施局部措施l通常用来消除变差的特殊原因l通常由与过程直接相关的人员实施l大约可纠正15%的过程问题对系统采取措施l通常用来消除变差的普通原因l几乎总是要求管理措施，以便纠正l大约可纠正85%的 过程问题33 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statis

27、tical Process ControlStatistical Process Control过程控制和过程能力过程控制和过程能力过程控制系统的目标是对影响过程的措施作出经济合理的决定，平衡不需控制是采取措施（过度控制或擅自改变）和需要控制是未采取措施（控制不足）的后果。过程在统计控制下运行仅存在造成变差的普通原因，对特殊原因采取适当措施（或是消灭，或是如果有用，永久保留它们）。34 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistic

28、al Process ControlStatistical Process Control35 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control顾客允许制造商运行一个类过程：顾客允许制造商运行一个类过程：顾客对规范要求之间的变差不敏感；顾客对规范要求之间的变差不敏感；对特殊原因采取措施所发生的成本比任何所有顾客得对特殊原因采取措施所发生的成本比任

29、何所有顾客得到的利益大，因成本原因允许存在特殊原因包括刀具磨到的利益大，因成本原因允许存在特殊原因包括刀具磨损、刀具重磨、周期的（季节的）变化；损、刀具重磨、周期的（季节的）变化；特殊原因已被查明，经记录表明具有稳定性和可预见特殊原因已被查明，经记录表明具有稳定性和可预见性。性。36 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control顾客会有以下要

30、求：顾客会有以下要求：该过程是成熟的；该过程是成熟的；允许存在的特殊原因在已知一段时间内表现出产允许存在的特殊原因在已知一段时间内表现出产生稳定的结果；生稳定的结果；过程控制计划有效运行，可确保所有的过程输出过程控制计划有效运行，可确保所有的过程输出符合规范并能防止出现别的特殊原因或与允许存在符合规范并能防止出现别的特殊原因或与允许存在的特殊原因不稳定的其他原因。的特殊原因不稳定的其他原因。37 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Stat

31、istical Process ControlStatistical Process Control说明：说明：汽车行业规定汽车行业规定“一个过程被证明处于统计控制一个过程被证明处于统计控制状态后才计算起过程能力。状态后才计算起过程能力。”短期能力短期能力验证首批产品、机器能力研究验证首批产品、机器能力研究 长期能力长期能力验证过程的性能验证过程的性能38 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process Cont

32、rolStatistical Process Control39 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control40 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Proces

33、s ControlStatistical Process Control控制图的选择控制图的选择41 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control统计过程控制统计过程控制计量型控制图计量型控制图lX-R均值与极差图lX-S均值与标准差图lX-R中位数图lX-MR单值和移动极差图计数型控制图计数型控制图lp不合格品率图lnp不合格品数图lc不合

34、格数图lu单位产品不合格数图42 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control计量性数据控制图43 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process Contro

35、lStatistical Process Control不精密不精密精精 密密不准确不准确准准 确确44 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control45 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training For

36、m Statistical Process ControlStatistical Process Control注意：注意：计量型控制图所有计算公式和分析原则计量型控制图所有计算公式和分析原则的基础是数据是成正态分布的的基础是数据是成正态分布的。因此，在运用控制图前必须确定数据本因此，在运用控制图前必须确定数据本身是否成正态分布，可以使用直方图或正态身是否成正态分布，可以使用直方图或正态概率纸来分析数据分布状况。概率纸来分析数据分布状况。46 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Int

37、ernal Training Form Statistical Process ControlStatistical Process Control案例案例47 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control48 Copy right reserved by TV Rheinland Academy Shanghai Edited by Ton

38、y Hung.Internal Training Form Statistical Process ControlStatistical Process Control49 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control均值和极差图使用均值和极差图使用50 Copy right reserved by TV Rheinland Academy

39、Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlA 收集数据收集数据A.1 选择子组大小、频度和数据选择子组大小、频度和数据a子组大小在过程初始研究时，子组一般由4到5件连续生产的产品组合，仅代表单一刀具、冲头、模腔等生产出的零件（即一个单一的过程流）b子组频率在过程的初期研究时，通常是连续进行分组或很短的时间间隔进行分组，以便检查过程在很短时间间隔内是否有其他不稳定因素存在。当证明过程处于稳定状态，子组间的时间间隔可以增加。51

40、Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlC 子组数的大小子组数的大小 l 从过程角度看，收集越多的子组可以确保变差的主要原因有机会出现（一般情况：包括100或更多单值读数的25或更多子组）。l如果过程已稳定，则可以得到过程的位置和分布宽度的有效 的估计值。l 在有些情况下，可以利用现有的数据来加速这个 第一阶段的研究。但只有他们是最近的，

41、并且对建立子组的基础很清楚的情况下才能使用。52 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control53 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process Contro

42、lStatistical Process ControlA.2 建立控制图及记录原始数据建立控制图及记录原始数据54 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control55 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Traini

43、ng Form Statistical Process ControlStatistical Process ControlA.4 选择控制图的刻度l对于X图，坐标上的刻度值的最大值与最小值之差应至少为子组均值（X）最大值与最小值差的2倍 l对于R图，刻度值应从最低值为0开始到最大值之间的差值为出事阶段所运道的最大极差（R）的2倍。A.5 将均值和极差值画到控制图上56 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Proce

44、ss ControlStatistical Process Control57 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlB.计算控制项计算控制项58 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form

45、 Statistical Process ControlStatistical Process Control59 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process ControlB.3 在控制图上作出平均值和极差值控制限的控制线在控制图上作出平均值和极差值控制限的控制线 在初始研究阶段，UCL、LCL被称为试验控制限。60 Copy right reserve

46、d by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control61 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control62 Copy right r

47、eserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control超出控制限的点超出控制限的点超出上控制限控制限计算错误或描点时描错；零件间的变化性或分布的宽度已经增大（即变坏），这种增大可以发生在某个时间点上，也可能是整个趋势的一部分；测量系统变化（例如，不同的检验员或量具）；测量系统没有适当的分辨力。有一点为于控制限之下控制限或描点错误；分布的宽度变小（即变好）；测量系统已改变（包括

48、数据编辑或变换）。63 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control64 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical

49、Process Controlb链现象链现象高于平均极差的链或上升链输出值的分布宽度增加，其原因可能是无规律（例如设备控制不正常或固定松动）或是由于过程中的某个要素变化（例如，使用新的不是很一致的原材料），需要纠正；测量系统改变（例如，新的检验员或量具）。底于平均极差的链，或下降链输出值分布宽度减小，应研究以便推广应用和改进过程；测量系统改变，这样会掩盖真实性能的变化。65 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Pro

50、cess ControlStatistical Process Control66 Copy right reserved by TV Rheinland Academy Shanghai Edited by Tony Hung.Internal Training Form Statistical Process ControlStatistical Process Control明显的非随机图形明显的非随机图形各点与的距离：一般地，大约的描点应落在控制限的中间的区域内，大约的点落在其外地区域。Lower SpecificationLimitUpperSpecificationLimitRed