An algorithm for selecting an optimal acceptance plan in quality control and auditing

This paper is concerned with the computational determination of minimum cost parameter values (sample size and acceptance number) for single sample. Bayesian acceptance plans in quality control when the prior distribution of lot quality and the sampling distribution are discrete. The cost surface of such problems can be described as a discrete, discontinuous function with numerous local optima. A two-stage optimization algorithm is developed for determining the optimal economic sampling plan. The effectiveness of this algorithm is systematically evaluated and shown to be very efficient both in terms of solution quality and computational time against existing solution procedures.