【作者】 李奔波；

【导师】 李传昭； 张旭梅；

【作者基本信息】 重庆大学 ， 技术经济及管理， 2006， 博士

【摘要】 自上世纪八十年代以来,随着市场竞争的加剧,现代企业所处的市场环境发生了深刻的变革,企业竞争越来越强调基于时间和客户需求的竞争。企业之间的竞争焦点也逐渐从价格因素向质量、柔性、快速交货等非价格因素转变,中小批量生产方式开始占有越来越重要的地位。这就向基于少品种、大批量的传统生产过程质量控制技术提出了挑战。本文针对多品种小批量生产方式的特点,综合运用管理学、概率论、数理统计等理论和方法,从以下几个方面对多品种小批量生产过程质量控制技术和方法进行了改善性研究。①提出了多品种小批量生产环境下工序能力的判定技术和方法。传统的工序能力评价采用工序能力指数的点估计值同推荐值进行比较,估计的正确性取决于样本数量的大小。在多品种小批量生产环境下,样本数目比较小,点估计不确定性将增加,导致工序能力指数的估计产生较大的偏差,可信度不高,风险较大。针对这种情况,本文首先对工序能力指数进行点估计,寻找其点估计的置信区间,将所得到的置信区间下限同推荐值进行比较,得出在一定置信度下能否完成加工的判定,而不是直接将点估计值同推荐值进行比较,从而得到该工序处于某个工序能力等级的判定。这样既增强了可信度,又降低了多品种小批量环境下工序能力判断的风险。为了提高实际的应用效果,对各种常用的工序能力指数( C P、C P上、C P下、C Pm和C Pw)的区间估计进行推导,得出了几种常用的工序能力指数在不同的样本容量下的置信区间,为多品种小批量生产环境下工序能力指数的应用提供了可能。同时,对基于成本控制的多品种小批量生产环境下的工序能力指数进行研究,指出工序能力指数不仅与样本容量有关,还与加工成本和次废品损失成本有关。②本文对多品种小批量生产环境下的单值控制图和中位数控制图进行了研究,分别提出两种控制图的控制界限的确定方法。对于多品种小批量生产环境下的单值控制图,构建了控制界限随着样本容量变化的单值控制图,给出了在不同样本容量下单值控制图控制界限的确定方法;对于多品种小批量生产环境下的中位数控制图,通过求出中位数的密度函数,求得在不同样本容量下标准化中位数近似服从正态分布的方差,再分别基于标准差和极差构造出新的统计量,得到中位数控制图动态的可变控制界限,从而实现中位数控制图对工序的控制,克服了将传统的控制图用于多品种小批量生产控制会导致“误发警报”概率增大的不足。③对多品种小批量生产环境下的多元控制图进行了研究。针对小批量生产环更多还原

【Abstract】 Since 1980s, with the sharpening of market competition, market environment in which enterprises existed in has been improved a lot. More and more enterprises realized the importance of competitive on time and satisfaction to customer demands. In the mean time, the focus of competition gradually changed from the price element to some non-price elements such as quality, flexibility and quick delivery. Consequently, the patterns of producing in middle and small batches gained a lot of attention, which is, of course, a great challenge to quality control techniques in the traditional producing process of small varieties and mass batches production.Based on the particularities of mass variety and small batches producing style, and also with the help of knowledge in Management, Probability, and Mathematical Statistics, the thesis will study the quality control techniques and methods in producing process of mass varieties and small batches production from the following respects.①Traditional processes capability evaluation compares the point estimate values and the recommended values of the Processes Capability Indices (PCI), and the accuracy of estimate depends on the sample size. In the environment of mass varieties and small batches production, the point estimates of PCI have large variations if the sample size is small. Therefore, instead of directly judging the grade of process capability, this thesis point estimate the PCI at first and find the confidence limits, then compare the lower confidence limits with recommended values to see if the process can be finished in certain confidence. In this way, it not only improves the curacy of estimate but also declines the risks of judging process capability in the environment of mass varieties and small batches production. To be practical, the confidence limits of some PCI ( C P、C P上、C P下、C Pmand C Pw) in several sample sizes are calculated.Meanwhile, the PCI of small batches production that based on cost control are studied and find that the PCI not only depends on sample size, but also relate to the cost of processing and the losing of substandard products.②Because of small batches and large variations, the probability of“False Alarm”may increase if the traditional control charts are used in mass varieties and small batches production. The thesis discusses the Single Control Chart and Median Control Chart separately, and gives the methods to fix the confidence limits of the two charts. As regards the single control chart in mass varieties and small batches production, the chart更多还原