| Summary: | 博士 === 國立雲林科技大學 === 工業工程與管理研究所博士班 === 101 === Because of the progress and development of high quality manufacturing processes, the traditional p-chart and np chart are no longer suitable for process monitoring involving the extremely low nonconforming fraction, p. A cumulative count of conforming (CCC) chart has been demonstrated to be useful in high quality manufacturing processes for monitoring the nonconforming fraction. When the quality characteristic is asymmetric and the one directional shifting is of interest, a one-sided control chart is more appropriate than a two-sided chart. Thus, we construct a one-sided CCC chart for high quality processes. Furthermore, some previous runs are considered in regular control procedures for establishing the conditional decision procedure of the one-sided CCC chart to improve its sensitivity. The nonconforming fraction parameter is assumed to be known when implementing a one-sided CCC chart. In this research, we also investigated the impact of estimated nonconforming fraction in a one-sided CCC chart. The run length distribution is derived as well as the conditional probability of a false alarm rate (CFAR), conditional average run length (CARL) and its standard deviation (CSDRL). Simulation results are conducted to evaluate the effect of estimator in a one-sided CCC chart. Finally, the proposed charting technique is demonstrated to monitor the medication error for a hospital pharmacy. The results show that the proposed method contributes to improving in the health care quality.
|
|---|
