Comparison and Evaluation of Performance for Standard Control Limits and Bootstrap Percentile Control Limits in $\bar{x}$ Control Chart

$\bar{x}$ 관리도의 표준관리한계와 부트스트랩 백분률 관리한계의 수행도 비교평가

  • 송서일 (동아대학교 산업공학과) ;
  • 이만웅 (동아대학교 산업공학과)
  • Published : 1999.11.01

Abstract

Statistical Process Control(SPC) which uses control charts is widely used to inspect and improve manufacturing process as a effective method. A parametric method is the most common in statistical process control. Shewhart chart was made under the assumption that measurements are independent and normal distribution. In practice, this assumption is often excluded, for example, in case of (equation omitted) chart, when the subgroup sample is small or correlation, it happens that measured data have bias or rejection of the normality test. A bootstrap method can be used in such a situation, which is calculated by resampling procedure without pre-distribution assumption. In this study, applying bootstrap percentile method to (equation omitted) chart, it is compared and evaluated standard process control limit with bootstrap percentile control limit. Also, under the normal and non-normal distributions, where parameter is 0.5, using computer simulation, it is compared standard parametric with bootstrap method which is used to decide process control limits in process quality.

Keywords