【摘要】：目前全球市場(chǎng)上不斷出現(xiàn)各種產(chǎn)品的召回事件,使得人們都越來(lái)越關(guān)心產(chǎn)品質(zhì)量,尤其是產(chǎn)品的來(lái)源及去向,而企業(yè)同時(shí)還注重提高問(wèn)題產(chǎn)品質(zhì)量原因分析的能力。根據(jù)電子元器件產(chǎn)品的特點(diǎn),本文研究了質(zhì)量追溯的關(guān)鍵技術(shù),即產(chǎn)品追蹤定位和問(wèn)題產(chǎn)品模糊診斷技術(shù)。產(chǎn)品追蹤主要用于搜索生產(chǎn)過(guò)程中所有可能造成質(zhì)量問(wèn)題的因子;模糊診斷主要對(duì)模糊現(xiàn)象進(jìn)行定量化診斷,從而得出造成質(zhì)量問(wèn)題的原因。但是目前在產(chǎn)品追蹤方面的研究主要集中在批次追蹤,缺少與質(zhì)量分析之間的聯(lián)系,使得其他重要信息缺失;在質(zhì)量原因分析方面,定量診斷的精度不夠高。因此質(zhì)量追溯要同時(shí)注重追蹤信息的齊全性和診斷精度的提升。針對(duì)電子元器件類產(chǎn)品追蹤定位,分析了一般追蹤算法追蹤的信息不完整并且追蹤信息雜亂無(wú)章的問(wèn)題,提出了本文的產(chǎn)品追蹤算法,包括前向追蹤算法和后向追蹤算法,并明確了兩者之間關(guān)系。該算法基于批次和約束,既注重批量生產(chǎn)的產(chǎn)品的特點(diǎn),也注重產(chǎn)品的生產(chǎn)工藝,并結(jié)合具有多元組節(jié)點(diǎn)的產(chǎn)品結(jié)構(gòu)樹(shù),彌補(bǔ)了一般追蹤算法的缺陷,同時(shí)可以得到多種退化型樹(shù)狀結(jié)構(gòu),為后續(xù)的質(zhì)量問(wèn)題模糊分析建模提供了依據(jù)。針對(duì)質(zhì)量問(wèn)題產(chǎn)品的原因分析,傳統(tǒng)的經(jīng)驗(yàn)判斷手段既無(wú)法定量化描述可能的原因,也無(wú)法提高定性分析的準(zhǔn)確度。在分析了其他行業(yè)運(yùn)用的診斷技術(shù)后,以提高診斷精確度為目的,提出了基于模糊診斷的質(zhì)量原因分析法。首先運(yùn)用統(tǒng)計(jì)法和退化型樹(shù)狀結(jié)構(gòu)對(duì)產(chǎn)品問(wèn)題征兆和原因進(jìn)行建模,提出了德?tīng)柗苾?yōu)序數(shù)法,并使用了數(shù)理統(tǒng)計(jì)和德?tīng)柗苾?yōu)序數(shù)法確定了各征兆和影響因子的隸屬度,隨后確定了模糊關(guān)系矩陣,最后采用了模型算子Ⅳ進(jìn)行診斷,并通過(guò)實(shí)例進(jìn)行驗(yàn)證,結(jié)果表明該方法不僅能診斷出一般方法診斷不出的結(jié)果,還具有更高的診斷精度。平臺(tái)信息化建設(shè)是企業(yè)提高管理水平的重要手段。基于產(chǎn)品追蹤和質(zhì)量問(wèn)題原因分析理論上的研究,設(shè)計(jì)與實(shí)現(xiàn)了軟件系統(tǒng),利用其強(qiáng)大的數(shù)據(jù)管理功能的優(yōu)勢(shì),為本文研究提供數(shù)據(jù)支撐和進(jìn)一步的理論驗(yàn)證。
[Abstract]:At present, there are a variety of product recalls in the global market, which makes people pay more and more attention to the quality of products, especially the origin and destination of products, while the enterprises also pay attention to improving the ability of cause analysis of the quality of the problem products. According to the characteristics of electronic component products, this paper studies the key technologies of quality traceability, that is, product tracking and positioning and fuzzy diagnosis of problematic products. Product tracking is mainly used to search all the factors that may cause quality problems in the production process, and fuzzy diagnosis is mainly used to quantitatively diagnose fuzzy phenomena, so as to obtain the causes of quality problems. However, the current research on product tracking mainly focuses on batch tracking, which is lack of connection with quality analysis, which makes other important information missing, and the accuracy of quantitative diagnosis is not high enough in quality cause analysis. Therefore, quality traceability should pay attention to the completeness of tracking information and the improvement of diagnostic accuracy at the same time. Aiming at the tracking location of electronic components, this paper analyzes the problem that the tracking information of the general tracking algorithm is incomplete and the tracking information is chaotic, and puts forward the product tracking algorithm in this paper, including the forward tracking algorithm and the backward tracking algorithm. The relationship between the two is clarified. Based on batch and constraint, the algorithm not only pays attention to the characteristics of batch production products, but also pays attention to the production process of products, and combines the product structure tree with multi-group nodes to make up for the defects of the general tracking algorithm. At the same time, a variety of degenerate tree structures can be obtained, which provides the basis for the subsequent fuzzy analysis modeling of quality problems. In view of the cause analysis of quality problems, the traditional means of empirical judgment can neither quantitatively describe the possible causes nor improve the accuracy of qualitative analysis. After analyzing the diagnostic techniques used in other industries, a quality cause analysis method based on fuzzy diagnosis is proposed in order to improve the diagnostic accuracy. Firstly, the statistical method and the degenerate tree structure are used to model the symptoms and causes of the product problem, and the Delphi superior ordinal number method is proposed, and the membership degree of each symptom and influence factor is determined by using mathematical statistics and Delphi optimal ordinal number method. Then the fuzzy relation matrix is determined and the model operator 鈪，

本文編號(hào)：2413423