废旧机床再制造过程质量控制方法研究

发布时间：2018-04-09 00:31

本文选题：再制造　切入点：质量控制　出处：《沈阳工业大学》2017年硕士论文

【摘要】：废旧机床再制造作为我国循环经济、绿色经济的重要组成部分之一,对于实现社会的可持续发展有着重大意义。而再制造产品质量是制约再制造企业发展的重要因素,同时,再制造系统存在许多的不确定性:废旧毛坯服役工况的不确定性、本身质量缺陷的不确定性以及组织结构的不确性,导致废旧机床再制造过程的质量控制较为困难,使得再制造产品的质量难以保证。基于此,以废旧机床再制造过程为研究对象,根据再制造过程的不确定性特点,对再制造过程质量异常在线监控、异常识别以及质量异常诊断与调整进行研究,主要研究内容如下:(1)针对毛坯质量分布不确定的问题,提出基于动态、非正态的EWMA控制图的方法。根据非参数方法中的Wilcoxon秩和检验的理论知识,应用秩统计量,获得与样本数据分布无关的统计量;在此基础上,通过不断移动控制图数据窗口来更新在线观测点,并利用得分函数获得动态的光滑参数,构建面向动态、非正态分布的再制造过程质量EWMA控制图,实现动态再制造过程质量的自适应监控。(2)针对再制造过程质量样本少、质量异常识别困难等问题,提出基于CS分解贝叶斯的主元分析方法。通过将CS分解贝叶斯空间估计的思想,特征子空间矩阵引入传统的PCA的方法,解决传统PCA需要大样本的问题,从而能实现在小样本条件下再制造过程质量异常识别。(3)针对再制造过程质量异常与引发因素之间的不确定性,采用基于粗糙集理论的再制造过程质量异常诊断与调整系统。利用粗糙集理论知识求得再制造过程质量影响因素相对再制造过程质量特征的重要度,再通过不同质量因素之间重要度大小,对质量异常进行溯源,找出引发质量异常的原因并进行调节,实现再制造过程的异常诊断与调整。(4)以废旧TPX6113镗床的导轨再制造过程为例,通过基于动态、非正态的EWMA控制图方法能灵敏地识别出再制造过程质量异常,出现质量异常后,基于改进的PCA质量异常识别模型能迅速识别出质量异常类型,再通过基于粗糙集的质量诊断模型对再制造质量异常诊断,诊断出质量异常原因并对过程异常进行调节,结果表明所提出的方法能够解决再制造过程质量控制问题,从而证明该方法的有效性和可行性。
[Abstract]:As one of the important parts of recycling economy and green economy in China, the remanufacture of waste machine tools is of great significance to the sustainable development of society.The quality of remanufacturing products is an important factor restricting the development of remanufacturing enterprises. At the same time, there are many uncertainties in the remanufacturing system: the uncertainty of service conditions of waste blank,The uncertainty of quality defects and the uncertainty of organization structure lead to the difficulty of quality control in the remanufacturing process of waste machine tools, which makes it difficult to guarantee the quality of remanufactured products.Based on this, taking the remanufacturing process of waste machine tool as the research object, according to the uncertainty characteristics of the remanufacturing process, this paper studies the on-line monitoring of the quality anomaly, the abnormal identification, the diagnosis and adjustment of the quality anomaly in the remanufacturing process.The main research contents are as follows: (1) to solve the problem of uncertain quality distribution of blank, a method based on dynamic and non-normal EWMA control chart is proposed.According to the theoretical knowledge of Wilcoxon rank sum test in nonparametric methods, the statistics independent of sample data distribution are obtained by using rank statistics, and the on-line observation points are updated by moving the control chart data window continuously.The dynamic smooth parameters are obtained by using the score function, and the EWMA control chart of the remanufacturing process quality oriented to dynamic and non-normal distribution is constructed, which realizes the self-adaptive monitoring of the quality of the dynamic remanufacturing process.This paper presents a principal component analysis method based on CS decomposition Bayes.Based on the idea of CS decomposition Bayesian space estimation, the characteristic subspace matrix is introduced into the traditional PCA method to solve the problem that traditional PCA needs large samples.So it can realize the recognition of remanufacturing process quality anomaly under the condition of small sample. Aiming at the uncertainty between the remanufacturing process quality anomaly and the trigger factor, a rough set theory based remanufacturing process quality anomaly diagnosis and adjustment system is adopted.By using rough set theory, the importance of the quality influencing factors in the remanufacturing process relative to the quality characteristics of the remanufacturing process is obtained, and then the source of the quality anomalies is traced through the importance degree between the different quality factors.Find out the cause of the abnormal quality and adjust it to realize the abnormal diagnosis and adjustment of the remanufacturing process. Take the remanufacturing process of the used TPX6113 boring machine as an example, based on the dynamic,The non-normal EWMA control chart method can sensitively identify the quality anomaly in the remanufacturing process. After the quality anomaly occurs, the improved PCA quality anomaly recognition model can quickly identify the type of the quality anomaly.Then, by using the quality diagnosis model based on rough set to diagnose the remanufacturing quality anomaly, the cause of the quality anomaly is diagnosed and the process anomaly is adjusted. The results show that the proposed method can solve the problem of the remanufacturing process quality control.Thus, the effectiveness and feasibility of the method are proved.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F426.4;F713.2;F273.2

【参考文献】