APPROACHES FOR PROBLEM DIAGNOSIS VIA STATISTICAL PROCESS CONTROL CHARTS

ABSTRACT Statistical process control charts aim   to detect out-of-control signals which indicate existence of special causes effecting the process. Once such a signal is detected, the interpretation of the signal, that is, discovering the actual causes behind the signal, rests upon the shoulders   of operators or engineers. Recently, some techniques have been developed   for making this interpretation process easier. This study presents an overview of such techniques in three categories: traditional one-variable control charts, artificial neural network applications and multivariate control charts.     Key Words: SPC   charts, out-of-control signal,   problem diagnosis