An Optimal Filter Design Approach to Statistical Process Control

• Autores: Daniel W. Apley, Chang-Ho Chin
• Localización: Journal of quality technology: A quarterly journal of methods applications and related topics, ISSN 0022-4065, Vol. 39, Nº. 2, 2007, págs. 93-117
• Idioma: inglés
• Enlaces
  • Texto completo
• Resumen

  • Many control charts are used to determine the output of a linear filter applied to process data. An alarm is sounded when the filter output falls outside a set of control limits. In this study, this concept is generalized by observing the linear filter in its most general time-invariant form. A strategy is provided for optimizing the filter coefficients to minimize the out-of-control average run length (ARL) while constraining the in-control ARL to a desired value. The optimal linear filters exhibit some characteristics of interest, particularly when the process data are autocorrelated. Often, they may outperform an optimally designed exponentially weighted moving average control chart.