• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.39 no.3
• /
• pp.10-17
• /
• 2016

Control Chart is a graph which dots the characteristic values of a process. It is the tool of statistical technique to keep a process in controlled condition. It is also used for investigating the state of a process. Therefore many companies have used Control Chart as the tool of statistical process control (SPC). Products from a production process represent accidental dispersion values around a certain reference value. Fluctuations cause of quality dispersion is classified as a chance cause and a assignable cause. Chance cause refers unmanageable practical cause such as operator proficiency differences, differences in work environment, etc. Assignable cause refers manageable cause which is possible to take actions to remove such as operator inattention, error of production equipment, etc. Traditionally ${\bar{x}}-R$ control chart or ${\bar{x}}-s$ control chart is used to find and remove the error cause. Traditional control chart is to determine whether the measured data are in control or not, and lets us to take action. On the other hand, RNELCC (Reflected Normal Expected Loss Control Chart) is a control chart which, even in controlled state, indicates the information of economic loss if a product is in inconsistent state with process target value. However, contaminated process can cause control line sensitive and cause problems with the detection capabilities of chart. Many studies on robust estimation using trimmed parameters have been conducted. We suggest robust RNELCC which used the idea of trimmed parameters with RNEL control chart. And we demonstrate effectiveness of new control chart by comparing with ARL value among traditional control chart, RNELCC and robust RNELCC.

• Journal of the Korean Data and Information Science Society
• /
• v.10 no.1
• /
• pp.233-241
• /
• 1999

Control chart is a very extensively used tool in testing whether a process is in a state of statistical control or not. In this paper, we propose a robust EWMA(exponentially weighted moving averages) control chart for variables, which is based on the Huber's M-estimator. The Huber's M-estimator is a well-known robust estimator in sense of distributional robustness. In the proposed chart, the estimation of the process deviation is modified to have a s table level and high power. To compare the performances of the proposed control chart with other charts, some Monte Carlo simulations we performed. The simulation results show that the robust EWMA control chart has good performance.

• Journal of Korean Society for Quality Management
• /
• v.40 no.3
• /
• pp.327-336
• /
• 2012

Purpose: In this study, we research the hurdle rate method to suggest the robust control chart for a contaminated process less vulnerable to fault values than existing control charts. Methods: We produce the results of p, ARL values to compare the performance of two control charts, $\bar{x}-s$ that has been used typically and TM-TS that is suggested by this paper. We implement the simulation focusing on three cases, change of deviation, mean and both of them. Results: We draw a conclusion that the TM-TS control chart has better efficiency than $\bar{x}-s$ control chart over the three cases. Conclusion: We insist that applying TM-TS control chart for a polluted process is more effective than $\bar{x}-s$ control chart.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.21 no.48
• /
• pp.65-71
• /
• 1998

Control chart is a very extensively used tool in testing whether a process is in a state of statistical control or not. In this paper, a robust control chart for variables is proposed, which is based on the Huber's M-estimator. The Huber's M-estimator is a well-known robust estimator in sense of distributional robustness. In the proposed chart, the estimation of the process deviation is modified to have a stable level and high power. To compare the performances of the proposed control chart with the classical (equation omitted), some Monte Carlo simulations are performed. The simulation results show that the robust control chart has good performance.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.26 no.3
• /
• pp.39-49
• /
• 2003

Statistical process cintrol is intended to assist operators of a stable system in monitoring whether a change has occurred in the process, and it uses several control charts as main tools. In design and use of control chart, it is rational that probability of false alarm is minimized in stable process and probability of detecting shifts is maximized in out-of-control. In this study, we establish bootstrap control limits for robust M-estimator chart by applying the bootstrap method, called resampling, which could not demand assumptions about pre-distribution when the process is skewed and/or the normality assumption is doubt. The results obtained in this study are summarized as follows : bootstrap M-estimator control chart is developed for applying bootstrap method to M-estimator chart, which is more robust to keep ARL when process contain contaminate quality characteristic.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.38 no.1
• /
• pp.15-20
• /
• 2015

Control charts are generally used for process control, but the role of traditional control charts have been limited in case of a non-contaminated process. Traditional ${\bar{x}}$-s control chart has 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 suggests modified ${\bar{x}}$-s control chart based on robust estimation. In this paper, we consider the trimmed mean of the sample means and the trimmed mean of the sample standard deviations. By comparing with ARL value, the responding results are decided. The comparison resultant results of traditional control chart and modified control chart are contrasted.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.39 no.1
• /
• pp.116-122
• /
• 2016

Recently, the production cycle in manufacturing process has been getting shorter and different types of product have been produced in the same process line. In this case, the control chart using coefficient of variation would be applicable to the process. The theory that random variables are located in the three times distance of the deviation from mean value is applicable to the control chart that monitor the process in the manufacturing line, when the data of process are changed by the type of normal distribution. It is possible to apply to the control chart of coefficient of variation too. ${\bar{x}}$, s estimates that taken in the coefficient of variation have just used all of the data, but the upper control limit, center line and lower control limit have been settled by the effect of abnormal values, so this control chart could be in trouble of detection ability of the assignable value. The purpose of this study was to present the robust control chart than coefficient of variation control chart in the normal process. To perform this research, the location parameter, ${\bar{x_{\alpha}}}$, $s_{\alpha}$ were used. The robust control chart was named Tim-CV control chart. The result of simulation were summarized as follows; First, P values, the probability to get away from control limit, in Trim-CV control chart were larger than CV control chart in the normal process. Second, ARL values, average run length, in Trim-CV control chart were smaller than CV control chart in the normal process. Particularly, the difference of performance of two control charts was so sure when the change of the process was getting to bigger. Therefore, the Trim-CV control chart proposed in this paper would be more efficient tool than CV control chart in small quantity batch production.

• Journal of Korean Society for Quality Management
• /
• v.29 no.1
• /
• pp.1-10
• /
• 2001

We consider the problem of detecting special variations in multivariate $T^2$-control chart when two or more multivariate outliers are present. Since a multivariate outlier may reflect slippage in mean, variance, or correlation, it can distort the sample mean vector and sample covariance matrix. Damaged sample mean vector and sample covariance matrix have difficulty in examining special variations clearly, An alternative to detection outliers or special variations is to use robust estimators of mean vector and covariance matrix that are less sensitive to extreme observations than are the standard estimators $\bar{x}$ and $\textbf{S}$. We applied popular minimum volume ellipsoid(MVE) and minimum covariance determinant(MCD) method to estimate mean vector and covariance matrix and compared its results with standard $T^2$-control chart using simulated multivariate data with outliers. We found that the modified $T^2$-control chart based on the above robust methods were more effective in detecting special variations clearly than the standard $T^2$-control chart.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.3
• /
• pp.729-738
• /
• 2015

In this paper, we compare the robustness to the assumption of normality of the single control charts to control the mean and variance simultaneously. The charts examined were semicircle control chart, max chart and MSE chart with Shewhart individuals control charts. Their in-control and out-of-control performance were studied by simulation combined with computation. We calculated false alarm rate to compare among single charts by changing subgroup size and shifting mean of quality characteristics. It turns out that max chart is more robust than any of the others if the process is in-control. In some cases max chart and MSE chart are more robust than others if the process is out-of-control.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1996.04a
• /
• pp.15-18
• /
• 1996

Shewhart control chart is a basic technique to monitor the state of a process. We observe observations of a group of size four or five in a rational way and plot some statistics (e.g., means and ranges) on the chart. When setting up the control chart, the control limits are calculated based on preliminary 20-40 samples, which were supposedly obtained from stable operating conditions. But it may be hard to believe, especially at the beginning of constructing the chart for the first time, whether the process is stable and hence all samples were generated under the homogeneous operating conditions. In this report we suggest a mechanism to obtain robust control limits under self-criticism. When outliers are present in the sample, we obtain tighter control limits and hence increase the sensitivity of the chart. Examples will be given via simulation study.