The Geometric Control Chart In Quality Management

In Statistical Quality Management, Statistical Process Control (SPC) is an importance aspect. In order to improve the manufacturing process in practice, we tried to make some research and analysis on this aspect especially on the Geometric Control Charts. In this article, we have found a new method and reached some meaning conclusions in the Quality Management.This article consists of two chapters. We give the summaries as follows:Chapter 1: As we all know, there are some common use control charts in Quality Management such as p or np chart. But they have also shortcoming in practice. The binomial p chart requires a binomial sample size n that is too large to be practical for many processes when p is as small as 0.01. The control chart based on the geometric distribution(g-type chart) has shown to be competitive with p or np chart for monitoring the proportion especially for application in high quality manufacturing environments. However when the cases with LCL=0 occur, plotting values beneath the LCL is impossible given the non-negative nature of units, such type of lower control limits no ability to detect rate increase. In fact, situation with a zero LCL is not uncommon. In this article we propose the g-type chart with asymmetric control limits. It is found that with the nonzero LCL, the power to detect rate increase hasmuch improved and the power to detect rate is nearly fixed, and the in-control average run length (ARL) is the almost the same as the symmetric control chart. In practice the quantitative results of the paper can be used to determine the control limits.Chapter 2: In general, if a point plots outside the control limits, we think the process is out of control. There have been numerous attempts to patch this deficiency in the geometric control chart with a zero LCL by adding "supplementary runs rules". The within-limit supplementary rules are that will lead to a signal if M successive points equal to the zero LCL M successive points equal to zero or one and if M successive points all plot beneath the center line (CL) even if they all remain inside the control limits. In this chapter, we gave examples to compare the three supplementary rules and reached some conclusions.