Design of an optimum adaptive control chart for attributes

Abstract

The classical (static) sampling strategies become less and less adequate for today's manufacturing practice because of their slow responses to process changes in the highly competitive industry society. Thus, it is desired to make better use of the inspection information and detect process shifts more promptly. An optimization scheme of designing a variable-sample-size attribute control chart is presented. The adjustment on sample size is based on the current process status observations for faster detection of any abnormality. As a result, a new adaptive-sample-size control chart for attributes, the VSS-np chart is proposed, using an optimization scheme. Comparison studies show that its performance characteristics is significantly better than traditional Shewhart control charts for attributes, since the design parameters are adaptive and obtained through the optimization procedures aimed at getting optimum average run length (ARL) properties.