APPLYING AN AR(1) CUSUM CONTROL CHART TO DATA FROM A CHEMICAL PROCESS Academic Article

abstract

• Control charts for variables are an important component of the quality program for many manufacturing processes. This article will present the application of a control chart to a physical property of a chemical in the outflow of a distillation column. This set of data is a representation of a large class of manufacturing processes in which the samples are not independent over time. The presence of serial correlation complicates data analysis and application of control charts due to its adverse effect on the run-length performance of traditional control charts. Techniques are presented to identify the presence of serial correlation in the data and identify appropriate alternative models for the data. In this case study, the first-order autoregressive model is used. Based on the first-order autoregressive model, parameters are identified and a control chart is developed for the data that are robust to the departures from independent samples (over time).

authors

• Longnecker, Michael

published proceedings

• Quality Engineering

author list (cited authors)

• Timmer, D. H., Pignatiello, J. J., & Longnecker, M. T.

citation count

• 5

complete list of authors

• Timmer, Douglas H||Pignatiello, Joseph J||Longnecker, Michael T

publication date

• February 2001

publisher

• Taylor & Francis  Publisher

published in

• Quality Engineering  Journal

Digital Object Identifier (DOI)

• 10.1080/08982110108918629

start page

• 107

end page

• 114

volume

• 13

issue

• 1

URL

• http%3A%2F%2Fdx.doi.org%2F10.1080%2F08982110108918629