本文關(guān)鍵詞：可變抽樣區(qū)間ARMA控制圖經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì)　出處：《鄭州航空工業(yè)管理學(xué)院》2017年碩士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì) 可變抽樣區(qū)間 ARMA控制圖 馬爾科夫鏈

【摘要】：控制圖是監(jiān)控質(zhì)量的重要工具,然而在自相關(guān)環(huán)境下,傳統(tǒng)控制圖會(huì)出現(xiàn)誤警報(bào)過于頻繁、系統(tǒng)不穩(wěn)定的缺陷,此時(shí)需要使用自相關(guān)控制圖如ARMA(Autoregressive Moving Average)等控制圖來監(jiān)控。為了盡可能的降低成本,減少企業(yè)使用控制圖的阻力,很多學(xué)者對(duì)控制圖進(jìn)行經(jīng)濟(jì)設(shè)計(jì)。但是控制圖的經(jīng)濟(jì)設(shè)計(jì)僅考慮了控制圖的經(jīng)濟(jì)性能,控制圖的統(tǒng)計(jì)性能大大減弱。因此,如何在保證質(zhì)量的前提下盡可能降低成本是一個(gè)值得研究的課題�？刂茍D的經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì)就是在保證控制圖監(jiān)控質(zhì)量的情況下盡可能的降低成本的一個(gè)方法。首先介紹了ARMA控制圖,建立了經(jīng)濟(jì)設(shè)計(jì)模型,以降低控制圖的使用成本,并在該模型的基礎(chǔ)上進(jìn)行改進(jìn),建立了考慮質(zhì)量損失函數(shù)的ARMA控制圖經(jīng)濟(jì)設(shè)計(jì)。其次,對(duì)可變抽樣區(qū)間ARMA控制圖進(jìn)行經(jīng)濟(jì)設(shè)計(jì),并與固定抽樣區(qū)間ARMA控制圖進(jìn)行比較得出,可變抽樣區(qū)間控制圖具有更低的成本,能夠更快的檢測(cè)出特殊原因的發(fā)生。最后對(duì)可變抽樣區(qū)間ARMA控制圖進(jìn)行經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì)。與相同條件的經(jīng)濟(jì)設(shè)計(jì)模型相比,經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì)模型的成本會(huì)小幅增加,但是控制圖的統(tǒng)計(jì)性能大大提高。本文層層遞進(jìn)的建立可變抽樣區(qū)間ARMA控制圖經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì)模型,并加入案例分析和求解,為企業(yè)提升產(chǎn)品質(zhì)量,降低生產(chǎn)成本提供了有效的途徑和理論支持。ARMA控制圖平均運(yùn)行長(zhǎng)度(Average Run Length,ARL)的馬爾可夫鏈法計(jì)算是一個(gè)難題。本文構(gòu)建滿足馬爾可夫鏈性質(zhì)的三維向量,計(jì)算狀態(tài)概率矩陣,進(jìn)而計(jì)算控制圖的ARL和平均報(bào)警時(shí)間(Average Time to Signal,ATS)。建立了可變抽樣區(qū)間ARMA控制圖經(jīng)濟(jì)統(tǒng)計(jì)設(shè)計(jì)模型,與ARMA控制圖經(jīng)濟(jì)設(shè)計(jì)相比,改模型的成本更低,統(tǒng)計(jì)性能更好。
[Abstract]:The control chart is an important tool for monitoring the quality, but in the autocorrelation environment, the traditional control chart will appear false alarm is too frequent, the instability of the system defects, the need to use autocorrelationcontrolcharts such as ARMA (Autoregressive Moving Average) control chart to monitor. In order to reduce the cost as far as possible, reduce the enterprise use control chart the resistance of many scholars to conduct economic design of control charts. But the economic design of control charts only consider the economic performance of control charts, statistical performance control chart is greatly reduced. Therefore, how to ensure the quality of the premise as far as possible to reduce costs is a topic worthy of study. The economic statistical design of control chart is in to ensure the quality control charts to monitor situations as much as possible to reduce a cost. First introduced the ARMA control chart, established the model of economic design, in order to reduce the use of control charts into This, and improved on the basis of the model, establish a control chart of economic design quality loss function of ARMA. Secondly, the variable sampling interval ARMA control chart and economic design, compared with the fixed sampling interval ARMA control chart with variable sampling interval control chart has lower cost, faster detection a special cause. Finally, the economic statistical design of variable sampling interval ARMA control chart. Compared with the economic design model and the same condition, the statistical design model of economic costs will increase slightly, but the statistical performance of control chart is greatly improved. A variable sampling interval control chart ARMA economic statistical model in this thesis and progressive layers. With case analysis and solution for enterprises to improve product quality, to provide the effective way and theoretical support for.ARMA control chart for average length to reduce the production cost ( Average Run Length, ARL) of the Markov chain method is a difficult problem. This paper constructed a three-dimensional vector Markov chain properties, calculation of state probability matrix, then calculate the average control chart and ARL (Average Time to Signal the alarm time, ATS). A variable sampling interval control chart ARMA economic statistical design model. Compared with the economic design of ARMA control model, lower cost, better statistical performance.

【學(xué)位授予單位】：鄭州航空工業(yè)管理學(xué)院
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：F224;F222

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 朱正森;;兩類控制圖不能混合一談[J];輕型汽車技術(shù);2002年08期

2 吳海英;合成控制圖[J];武漢理工大學(xué)學(xué)報(bào);2004年12期

3 王敏華;周娟;沈丹;;累積和控制圖與休哈特控制圖聯(lián)合應(yīng)用方法[J];湖北工業(yè)大學(xué)學(xué)報(bào);2008年05期

4 王成斌;回歸控制圖[J];數(shù)理統(tǒng)計(jì)與管理;1988年01期

5 嚴(yán)義榮,丁德儀;控制圖技術(shù)在工程地質(zhì)測(cè)試數(shù)據(jù)中的應(yīng)用[J];水文地質(zhì)工程地質(zhì);1991年04期

6 李躍波;合成控制圖[J];數(shù)理統(tǒng)計(jì)與管理;2001年06期

7 莊常陵;控制圖犯第Ⅰ/Ⅱ類錯(cuò)誤分析[J];長(zhǎng)春師范學(xué)院學(xué)報(bào);2001年02期

8 張維銘;可變抽樣區(qū)間的單邊控制圖[J];數(shù)理統(tǒng)計(jì)與管理;2002年06期

9 黃贏;控制圖在檢測(cè)實(shí)驗(yàn)室質(zhì)量控制中的應(yīng)用[J];現(xiàn)代測(cè)量與實(shí)驗(yàn)室管理;2002年04期

10 張公緒,孫靜;統(tǒng)計(jì)過程控制與診斷 第四講 ，

本文編號(hào)：1371909