Abordagem econômica-estatística para planejamento de controle on-line por atributo

Abstract

The procedure for online control by attribute consists of inspecting a single item at every m items produced (m>=2). On each inspection, it is decided whether the fraction of the produced non conforming items increased. If the inspected item is classified as nonconforming, the production process is adjusted so that the conforming fraction returns to its original status. A generalization observed in the literature is to consider missclassification errors, variable inspection interval and repetitive testing of the inspected item. In general, the authors offer an economic approach to establish the search procedure for the design control system, by minimizing the control system expecteded cost. This study presents a short review about the economical-statistical design of online process control by attribute. The mechanisms of multiobjective optimization by genetic algorithms is discussed and used to search the set of solutions with the optimal parameters that minimizes the expected cost and fraction of nonconforming itens sent to the market. The procedure is illustrated by numerical examples. This method uses the properties of an ergodic Markov chain to obtain the expression for the average cost of this control system.