【文章內(nèi)容簡(jiǎn)介】 Edited by Tony Hung. Internal Training Form169。 Statistical Process Control49 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control均值和極差圖使用50 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process ControlA． 收集數(shù)據(jù) 選擇子組大小、頻度和數(shù)據(jù)a．子組大小在過程初始研究時(shí)，子組一般由 4到 5件連續(xù)生產(chǎn)的產(chǎn)品組合，僅代表單一刀具、沖頭、模腔等生產(chǎn)出的零件（即一個(gè)單一的過程流）b．子組頻率在過程的初期研究時(shí)，通常是連續(xù)進(jìn)行分組或很短的時(shí)間間隔進(jìn)行分組，以便檢查過程在很短時(shí)間間隔內(nèi)是否有其他不穩(wěn)定因素存在。當(dāng)證明過程處于穩(wěn)定狀態(tài)，子組間的時(shí)間間隔可以增加。51 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process ControlC． 子組數(shù)的大小 l 從過程角度看，收集越多的子組可以確保變差的主要原因有機(jī)會(huì)出現(xiàn)（一般情況：包括 100或更多單值讀數(shù)的 25或更多子組）。 l 如果過程已穩(wěn)定，則可以得到過程的位置和分布寬度的有效 的估計(jì)值。 l 在有些情況下，可以利用現(xiàn)有的數(shù)據(jù)來加速這個(gè) 第一階段的研究。但只有他們是最近的，并且對(duì)建立子組的基礎(chǔ)很清楚的情況下才能使用。52 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control53 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control 建立控制圖及記錄原始數(shù)據(jù)54 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control55 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control 選擇控制圖的刻度l對(duì)于 X圖，坐標(biāo)上的刻度值的最大值與最小值之差應(yīng)至少為子組均值（ X）最大值與最小值差的 2倍 l 對(duì)于 R圖，刻度值應(yīng)從最低值為 0開始到 最大值之間的差值為出事階段所運(yùn)道的最大極差（ R）的 2倍。 將均值和極差值畫到控制圖上56 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control57 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process ControlB. 計(jì)算控制項(xiàng)58 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control59 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control 在控制圖上作出平均值和極差值控制限的控制線 在初始研究階段， UCL、 LCL被稱為試驗(yàn)控制限。60 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control61 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control62 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Controlａ．超出控制限的點(diǎn)超出上控制限控制限計(jì)算錯(cuò)誤或描點(diǎn)時(shí)描錯(cuò)；零件間的變化性或分布的寬度已經(jīng)增大（即變壞），這種增大可以發(fā)生在某個(gè)時(shí)間點(diǎn)上，也可能是整個(gè)趨勢(shì)的一部分；測(cè)量系統(tǒng)變化（例如，不同的檢驗(yàn)員或量具）；測(cè)量系統(tǒng)沒有適當(dāng)?shù)姆直媪?。有一點(diǎn)為于控制限之下控制限或描點(diǎn)錯(cuò)誤；分布的寬度變?。醋兒茫粶y(cè)量系統(tǒng)已改變（包括數(shù)據(jù)編輯或變換）。63 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control64 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Controlb．鏈現(xiàn)象高于平均極差的鏈或上升鏈輸出值的分布寬度增加，其原因可能是無規(guī)律（例如設(shè)備控制不正?；蚬潭ㄋ蓜?dòng)）或是由于過程中的某個(gè)要素變化（例如，使用新的不是很一致的原材料），需要糾正；測(cè)量系統(tǒng)改變（例如，新的檢驗(yàn)員或量具）。底于平均極差的鏈，或下降鏈輸出值分布寬度減小，應(yīng)研究以便推廣應(yīng)用和改進(jìn)過程；測(cè)量系統(tǒng)改變，這樣會(huì)掩蓋真實(shí)性能的變化。65 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control66 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process ControlＣ．明顯的非隨機(jī)圖形各點(diǎn)與Ｒ的距離：一般地，大約２／３的描點(diǎn)應(yīng)落在控制限的中間１／３的區(qū)域內(nèi)，大約１／３的點(diǎn)落在其外地／３區(qū)域。Lower SpecificationLimitUpperSpecificationLimitReducing Variation is Clearly the Key to ImprovingProcess Capability 3s1/367 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control顯著多于２／３以上的描點(diǎn)落在離Ｒ很近之處（對(duì)于 25個(gè)子組，如果超過 90％的點(diǎn)落在控制限１／３的區(qū)域） 控制限或描點(diǎn)已計(jì)算錯(cuò)或描錯(cuò)； 過程或取樣方法被分層；每個(gè)子組系統(tǒng)化包括了從兩個(gè)或多個(gè)具有完全不同的過程均值的過程流（例如，用幾根測(cè)量軸每一軸測(cè)一個(gè)數(shù)）； 數(shù)據(jù)已經(jīng)過編輯（極差與均值相差甚遠(yuǎn)的幾個(gè)子組被更改或刪除）。68 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control分組情況69 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control70 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control71 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Control72 Copy right reserved by T220。V Rheinland Academy Shanghai Edited by Tony Hung. Internal Training Form169。 Statistical Process Contro