• 지질공학
• /
• 제24권2호
• /
• pp.191-199
• /
• 2014

We describe a method for detecting slope instability in flume tests using pore pressure and water content data in conjunction with a statistical control chart analysis. Specifically, we conducted univariate statistical analysis on x-MR control chart data (pore pressure and water content) collected at several points along the flume slope, which we separated into three parts: upper, middle, and lower. To assess our results in the context of landslide forecasting and warning systems, we applied control limit lines at $1{\sigma}$, $2{\sigma}$, and $3{\sigma}$ levels of uncertainty. In doing so, we observed that dispersion time varies depending on the control limit line used. Moreover, the detection of instabilities is highly dependent on the position and type of sensor. Our findings indicate that different characteristics of the data on various factors predict slope failure differently and these characteristics can be identified by univariate statistical analysis. Therefore, we suggest that a univariate statistical approach is an effective method for the early detection of slope instability.

• 품질경영학회지
• /
• 제40권4호
• /
• pp.467-480
• /
• 2012

Purpose: This paper introduces new 2-of-3 main and supplementary runs rules to increase the performance of the classical $\bar{X}$ control chart for detecting small process shifts. Methods: The proposed runs rules are compared with other competitive runs rules by numerical experiments. Nonlinear optimization problem to minimize the out-of-control ARL at a specified shift of process mean for determining action and warning limits at a time is formulated and a procedure to find two limits is illustrated with a numerical example. Results: The proposed 2-of-3 main and supplementary runs rules demonstrate an improved performance over other runs rules in detecting a sudden shift of process mean by simultaneous changes of mean and standard deviation. Conclusion: To increase the performance in the detection of small to moderate shifts, the proposed runs rules will be used with $\bar{X}$ control charts.

• Industrial Engineering and Management Systems
• /
• 제6권1호
• /
• pp.11-21
• /
• 2007

The PDCA (Plan, Do, Check and Act) cycle is often used in the field of quality management. Recently, business environments have become more competitive, and the due time of products has shortened. In a short production run process, to increase efficiency of management, the necessity for distinguishing the PDCA design that starts with PLAN and the CAPD design that starts with CHECK has been clarified. Starting from Duncan (1956), there have been a number of papers dealing with the economic design of control charts from the viewpoint of production run. Some authors (Gibra, 1971; Ladany and Bedi, 1976; etc.) have studied the economic design for finite-length runs; other authors (Crowder, 1992; Del Castillo and Montgomery, 1996; etc.) have studied the economic design for short runs. However, neither the PDCA nor the CAPD design of control charts has been considered. In this paper, both the PDCA and CAPD designs of the $\bar{\x}$ chart are defined based on Del Castillo and Montgomery's design (1996), and their mathematical formulations are shown. Then from an economic viewpoint, the optimal values of the sample size per each sampling, control limits width, and the sampling interval of the two designs are studied. Finally, by numerically analyzing the relations between the key parameters and the total expected cost per unit time, the comparisons between the two designs are considered in detail.

• 품질경영학회지
• /
• 제37권2호
• /
• pp.46-57
• /
• 2009

The traditional approach to economic design of control charts is based on the assumption that a process is monitored using a performance variable. However, various types of automatic test equipments recently introduced as a part of factory automation usually measure surrogate variables instead of performance variables that are costly to measure. In this article we propose a model for economic design of a control chart which uses a surrogate variable that is highly correlated with the performance variable. The optimum values of the design parameters are determined by maximizing the total average income per cycle time. Numerical studies are performed to compare the proposed $\bar{X}$ control charts with the traditional model using the examples in Panagos et al. (1985).

• 응용통계연구
• /
• 제35권6호
• /
• pp.725-737
• /
• 2022

통계적 공정 모니터링에서 관리 상태일 때 품질 특성치의 모수값은 하나의 값으로 지정하는 경우가 대부분이다. 그러나 관리 상태로부터 공정 모수의 작은 변화는 실제적으로 크게 중요하지 않은 경우, 품질 특성치의 모수 영역은 관리 상태, 무관심, 그리고 이상 상태의 세 영역으로 구성될 수 있다. 이 논문에서는 3개의 모수 영역이 있는 공정에 적용할 수 있는 두 가지 지수가중 이동평균(exponentially weighted moving average; EWMA) 관리도 절차를 제안하고, 제안된 절차의 성능을 Shewhart 관리도 및 누적합(cumulative sum; CUSUM) 관리도와 비교하여 그 효율을 평가하였다.

• 산업경영시스템학회지
• /
• 제37권1호
• /
• pp.151-156
• /
• 2014

Control charts are generally used for process control, but the role of traditional control charts have been limited in case of a contaminated process. Traditional $\bar{x}$ control charts have not been activated well for such a problem because of trying to control processes as center line and control limits changed by the contaminated value. This paper is to propose robust $\bar{x}$ control charts which is considering a location parameter in order to respond to contaminated process. In this paper, we consider $\bar{x}_{\alpha}$, that is trimmed rate; typically ten percent rate is used. By comparing with p, ARL value, the responding results are decided. The comparison resultant results of proposed two control charts are shown and are well contrasted.

• Journal of the Korean Data and Information Science Society
• /
• 제26권4호
• /
• pp.1027-1034
• /
• 2015

A control chart is very useful in monitoring various production process. There are many situations in which the simultaneous control of two or more related quality variables is necessary. When the joint distribution of the process variables is multivariate normal, multivariate Shewhart control charts using the function of the maximum likelihood estimator for monitoring the dispersion matrix are considered for the simultaneous monitoring of the dispersion matrix. The performances of the multivariate Shewhart control charts based on the proposed control statistic are evaluated in term of average run length (ARL). The performance is investigated in three cases, where the variances, covariances, and variances and covariances are changed respectively. The numerical results show that the performances of the proposed multivariate Shewhart control charts are not better than the control charts using the trace of the covariance matrix in the Jeong and Cho (2012) in terms of the ARLs.

• 품질경영학회지
• /
• 제44권3호
• /
• pp.587-600
• /
• 2016

Purpose: This paper evaluates the performance of the pre-control(PC), an alternative to statistical process control techniques and compares with a control chart considering the tolerance of process. Methods: The previous studies for PC have drawbacks that PC with two linked stages, qualification and running, are discussed separately and independently. Hence this paper analyzes the performance of PC by integrating two stages. Results: Average outgoing quality limits to grasp the outcome of PC are provided by computational results for two process capability indexes, $C_p$ and $C_{pk}$ and the usefulness of PC from comparative experiments with modified control charts is commented. Conclusion: Helpful guidelines for quality managers to apply PC in practice and areas of process for PC to be more benefit are presented.

• 한국컴퓨터정보학회논문지
• /
• 제25권12호
• /
• pp.73-82
• /
• 2020

반도체나 석유화학 공정과 같이 프로세스 중심의 장치 산업에서는 흔히 관측된 자료들 사이에 자기상관(Autocorrelation)이 존재하는데, 이러한 공정에 기존의 SPC(Statistical process control)를 적용하는 경우 공정의 평균 변화를 효과적으로 검출하지 못하는 문제가 발생할 수 있다. 본 논문에서는 특정 시계열 모형을 따르는 공정자료에 일정한 크기의 평균 변화가 발생할 때, 잔차는 시간의 흐름에 따라 그 평균이 달라지게 되는데, ARMA(1,1) 과정을 중심으로 평균의 변화 패턴을 소개하고, 이 결과를 바탕으로 공정의 작은 평균 변화를 검출할 수 있는 CUSUM(Cumulative sum) 관리도의 공정 자료가 갖는 시계열 모형의 형태와 관심 있는 공정 평균 변화의 폭을 고려하여 CUSUM 관리도의 설계 과정에서 필요한 참고값이 적절히 선택되어 사용되어야 함을 모의실험을 통해 확인하였다.

• 한국항해항만학회지
• /
• 제29권3호
• /
• pp.269-274
• /
• 2005

현재 국내에서 활발히 이루어지고 있는 GPS 정밀측지측량은 GPS 관측성과가 관측값의 통계적 검증없이 이루어지고 있단 문제점이 있다. 그리고 육지와 해상의 통합지리정보시스템구축을 위한 수치지형도와 수치해도의 통합투영의 필요성이 요구되고 있다. 본 연구를 통하여 GPS 망조정에 사용되는 3점의 기준선에 대한 기하학적인 강도에 따른 정밀도 분석과 분산계수의 변화에 따른 GPS 관측 값들의 정밀도 분석방법으로 GPS 관측값의 정밀도 향상을 위한 방법을 제시하였다. 또한, GPS 관측값의 정밀도를 향상시키는 방법과 수치지도와 전자해도의 통합투영으로 GPS 기준점을 이용한 수지지형도와 수치해도의 경계선 오차량을 비교 검토하여 수치지형도와 수치해도의 통합관리의 방향을 제시하였다. 그리고 이렇게 산출된 GPS 관측결과를 이용하여 EDM 측량결과를 산출하고 수치지형도와 수치해도를 이 정계값들로 맞추었을 때 전체적인 경계점 오차량을 비교 검토하였다.