ALGORITMO PARA MINIMIZAÇÃO DO TAMANHO DE AMOSTRA EM PLANOS DE INSPEÇÃO POR AMOSTRAGEM POR ATRIBUTO

Abstract

The acceptance sampling plans for attribute have been largely used in the industrial sector and, they have assumed an important role for the quality control of raw material, semi-finished and finished products. When developing a sampling plan, the challenge is to obtain a small sample size as well as protect both the producer and the consumer of the involved risks. In this way, this paper was aimed at the development of an algorithm to identify a sampling plan, which provides the smallest sample size and that guarantees the restrictions of the producer and consumer’s risks simultaneously. The method employed was the modelling method. In order to illustrate the propose algorithm, a problem situation was simulated. As a result, an algorithm was obtained that is: easily implemented, for instance, it was implemented in software R; flexible, it allows us to insert values for the producer and consumer’s risks; an objective, because it provides a unique and optimal solution. If a sampling plan is developed by using the proposed algorithm, it will result in the smallest possible sample size. Therefore, the plan brings benefits like short inspection time and small quantity of manipulated items.