A double neural network approach for the automated detection of quality control chart patterns

Abstract

A control chart is one of the key tools in statistical process control. The exhibited pattern on a control chart indicates either a process is in control or out of control. The control chart patterns are classified to natural and unnatural patterns. The presence of unnatural patterns is evidence that the process is out of control. This paper proposes an artificial neural network algorithm to detect and identify any of the five basic control chart patterns; namely, natural, upward shift, downward shift, upward trend, and downward trend. This identification is in addition to the traditional statistical detection of sequential data runs. It is assumed that a process starts in control (has natural pattern) and then may undergo only under one out-of-control pattern at a time. The performance of the proposed algorithm was evaluated by measuring the probability of success in identifying the five basic patterns accurately and comparing these results with previous research. The comparison showed that the proposed algorithm is comparable if not superior.