• Journal of the Korean Operations Research and Management Science Society
• /
• v.36 no.1
• /
• pp.39-55
• /
• 2011

Process monitoring has been emphasized for the monitoring of complex system such as chemical processing industries to achieve the efficiency enhancement, quality management, safety improvement. Recently, ICA (Independent Component Analysis) based MSPC (Multivariate Statistical Process Control) was widely used in process monitoring approaches. Moreover, DICA (Dynamic ICA) has been introduced to consider the system dynamics. However, the existing approaches show the limitation that their performances are strongly dependent on the statistical distributions of control variables. To improve the limitation, we propose a novel approach for process monitoring by integrating DICA and LOF (Local Outlier Factor). In this paper, we aim to improve the fault detection rate with the proposed method. LOF detects local outliers by using density of surrounding space so that its performance is regardless of data distribution. Therefore, the proposed method not only can consider the system dynamics but can also assure robust performance regardless of the statistical distributions of control variables. Comparison experiments were conducted on the widely used benchmark dataset, Tennessee Eastman process (TE process), and showed the improved performance than existing approaches.

• Industrial Engineering and Management Systems
• /
• v.14 no.4
• /
• pp.392-400
• /
• 2015

Monitoring of plasma etch processes for fault detection is one of the hallmark procedures in semiconductor manufacturing. Optical emission spectroscopy (OES) has been considered as a gold standard for modeling plasma etching processes for on-line diagnosis and monitoring. However, statistical quantitative methods for processing the OES data are still lacking. There is an urgent need for a statistical quantitative method to deal with high-dimensional OES data for improving the quality of etched wafers. Therefore, we propose a robust relevance vector machine (RRVM) for regression with statistical quantitative features for modeling etch rate and uniformity in plasma etch processes by using OES data. For effectively dealing with the OES data complexity, we identify seven statistical features for extraction from raw OES data by reducing the data dimensionality. The experimental results demonstrate that the proposed approach is more suitable for high-accuracy monitoring of plasma etch responses obtained from OES.

• Communications for Statistical Applications and Methods
• /
• v.7 no.3
• /
• pp.773-788
• /
• 2000

Statistical process control (SPC) and engineering process control (EPC) are based on different strategies for processes quality improvement. SPC reduces process variability by detecting and eliminating special causes of process variation. while EPC reduces process variability by adjusting compensatory variables to keep the quality variable close to target. Recently there has been needs for a process control proceduce which combines the tow strategies. This paper considers a combined scheme which simultaneously applies SPC and EPC techniques to reduce the variation of a process. The process model under consideration is an integrated moving average(IMA) process with a step shift. The EPC part of the scheme adjusts the process back to target at every fixed monitoring intervals, which is referred to a repeated adjustment scheme. The SPC part of the scheme uses an exponentially weighted moving average(EWMA) of observed deviation from target to detect special causes. A Markov chain model is developed to relate the scheme's expected cost per unit time to the design parameters of he combined control scheme. The expected cost per unit time is composed of off-target cost, adjustment cost, monitoring cost, and false alarm cost.

• International Journal of Quality Innovation
• /
• v.1 no.1
• /
• pp.97-105
• /
• 2000

In this paper, a two-section method for measuring is introduced and the variation sources of measurement process are analysed. Measuring is a special process in general process. Various variation source must be firstly decomposed so that the statistical distribution law of measuring process can be established, and then implement monitoring control of the measuring process. A special method to obtain the measuring variation is discussed, and a monitoring control technique for measuring process is studied based statistical distribution. Towards the end, we briefly introduce software design for the analysis and control of a measurement process.

• Smart Structures and Systems
• /
• v.12 no.2
• /
• pp.181-207
• /
• 2013

In this paper, a statistical reference-free real-time damage detection methodology is proposed for detecting joint and member damage of truss bridge structures. For the statistical damage sensitive index (DSI), wavelet packet decomposition (WPD) in conjunction with the log likelihood ratio was suggested. A sensitivity test for selecting a wavelet packet that is most sensitive to damage level was conducted and determination of the level of decomposition was also described. Advantages of the proposed method for applications to real-time health monitoring systems were demonstrated by using the log likelihood ratios instead of likelihood ratios. A laboratory truss bridge structure instrumented with accelerometers and a shaker was used for experimental verification tests of the proposed methodology. The statistical reference-free real-time damage detection algorithm was successfully implemented and verified by detecting three damage types frequently observed in truss bridge structures - such as loss of bolts, loosening of bolts at multiple locations, sectional loss of members - without reference signals from pristine structure. The DSI based on WPD and the log likelihood ratio showed consistent and reliable results under different damage scenarios.

• Journal of Advanced Marine Engineering and Technology
• /
• v.24 no.4
• /
• pp.405-413
• /
• 2000

This paper present study on the application of sound pressure and vibration signals to detect the presence of defects in a rolling element bearing using a statistical analysis method. The well established statistical parameters such as the crest factor and the distribution of moments including kurtosis and skew are utilized in this study. In addition, other statistical parameters derived from the beta distribution function are also used. A comparison study on the performance of the different types of parameter used is also performed. The statistical analysis is used because of its simplicity and quick computation. Under ideal conditions, the statistical method can be used to identify the different types of defect present in the bearing. In addition, the results also reveal that there is no significant advantages in using the beta function parameters when compared to using kurtosis and the crest factor for detecting and identifying defects in rolling element bearings from both sound and vibration signals.

• Communications for Statistical Applications and Methods
• /
• v.21 no.1
• /
• pp.69-80
• /
• 2014

This paper considers the problem of monitoring the mean of a normally distributed process variable when the objective is to effectively detect both a sustained shift and a linear drift. The design and application of a generalized likelihood ratio (GLR) chart for simultaneously monitoring a sustained shift and a linear drift are evaluated. The GLR chart has the advantage that when we design this chart, we do not need to specify the size of the parameter change. The performance of the GLR chart is compared with that of other control charts, such as the standard cumulative sum (CUSUM) charts and the cumulative score (CUSCORE) charts. And we compare the proposed GLR chart with the GLR charts designed for monitoring only a sustained shift and for monitoring only a linear drift. Finally, we also compare the proposed GLR chart with the chart combinations. We show that the proposed GLR chart has better overall performance for a wide range of shift sizes and drift rates relative to other control charts, when a special cause produces a sustained shift and/or a linear drift in the process mean.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2007.11a
• /
• pp.1386-1390
• /
• 2007

This paper presents the equipment condition monitoring technology for the process or the equipment using statistical techniques. The equipment condition monitoring system consists of an empirical model to estimate the expected sensor values of process variables and a diagnose model to detect the abnormal condition and to identify the root source of the problem. The empirical model is constructed by the analysis of historic data. The diagnose model uses the sequential probability ratio test (SPRT) technique. The monitoring system was tested with real operating data acquired from the Reactor Coolant Pump Seal in the Nuclear Power Plant. It can detect the system degradation or failure at the early stage since it is able to catch the subtle deviation of process variables from normal condition.

• Communications for Statistical Applications and Methods
• /
• v.29 no.4
• /
• pp.487-498
• /
• 2022

Social network analysis (SNA) techniques have recently been developed to monitor and detect abnormal behaviors in social networks. As a useful tool for process monitoring, control charts are also useful for network monitoring. In this paper, the degree and closeness centrality measures, in which each has global and local perspectives, respectively, are applied to an exponentially weighted moving average (EWMA) chart and a multinomial cumulative sum (CUSUM) chart for monitoring undirected weighted networks. In general, EWMA charts monitor only one variable in a single chart, whereas multinomial CUSUM charts can monitor a categorical variable, in which several variables are transformed through classification rules, in a single chart. To monitor both degree centrality and closeness centrality simultaneously, we categorize them based on the average of each measure and then apply to the multinomial CUSUM chart. In this case, the global and local attributes of the network can be monitored simultaneously with a single chart. We also evaluate the performance of the proposed procedure through a simulation study.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.57 no.1
• /
• pp.1-9
• /
• 2015

In this study it was adopted on sea dike monitoring that the safety monitoring with statistical limits which was adapted usually on safety monitoring by measuring pressures, stresses or deformations. And also the hydraulic head loss rate change according to passage of time was calculated for the purpose of safety monitoring. Safety monitoring by setting the statistical limit on the measured pore water pressure graphs need to be supplemented with an additional method of monitoring because the difference between the rise and fall of the tide was irregular. Safety monitoring by the limits set from values predicted by linear regression and standard errors on the hydraulic head loss graph was not affected by irregularity of tide. But if the condition of an embankment is changed gradually and slowly, it will not be detected on the hydraulic head loss graph. The graph of hydraulic head loss rate for every 24 hours vs date showed clearly that the sea water blocking state was getting better or not even though it was changed gradually and slowly.