How to use EWMA to Achieve SPC and EPC Control
Adriano Mendonça Souza, UFSM, Santa Maria, RS, Brasil
ABSTRACT To control a process there are two methods, statistical process control (SPC) and engineering process control (EPC). Both methods utilize measurements from the process. EPC works with feedback control and uses measurements to prescribe changes to the process input intended to process outputs closer to their targets. SPC uses measurements to monitor a process and look for major changes, but does not prescribe control actions. Until recently there has been little overlap between these two approaches to process control. The goal of this work is to combine feedback control and SPC through EWMA statistic to reduce the variability of the process output, as measured by the mean squared deviation from the target. This tool will be used as control chart to monitoring the process and as estimator to forecast the next value the variable in study. In this way EWMA has a compromise between SPC and EPC. To make the adjustment through the feedback control will be utilized the proportional integral control and to reach this maim purpose will be discussed how to chose the best value for the smoothing constant in EWMA statistics and its applications in SPC and EPC. To show how this compromise could be achieved will be make one application in one univariate time series of temperature from "Cecrisa S.A. - Cerâmicas de Revestimentos", this way we have a framework to discuss the methodology available and make adjustments in the process in scientific basis.Key words: safety, reliability, and risk assessment, quality control
Publication Source: NDTISS '99 - International Symposium on NDT Contribution to the Infrastructure Safety Systems, Nov 1999 in Tores Brazil.
Publisher: Center of Tecnology, Federal University of Santa Maria (UFSM) - [Homepage]
© NDTnet
