PATTERN-RECOGNITION FOR STATISTICAL PROCESS-CONTROL CHARTS

Control charts are important statistical process control tools for det ermining whether a process is run in its intended mode or in the prese nce of unnatural patterns. Patterns displayed on control charts can pr ovide information about the process. This paper describes the developm ent of a pattern recognition system designed to detect and analyse var ious patterns that can occur on statistical quality control charts. Th e system looks not only for simple patterns, such as trend shift and s tratification, but also for superimposed patterns, such as trend + shi ft. The effect of noise associated with individual patterns is also an alysed. The benefits of the approach compared with the alternatives ar e discussed.