Average outgoing quality limit and lot tolerance percent defective indexed acceptance sampling plans using artificial neural networks

For maintaining quality at a target Average Outgoing Quality Limit (AOQL) and not below some quality value such as Lot Tolerance Percent Defective (LTPD) if we use the Dodge–Romig table, it offers limited flexibility to quality control engineers in designing sampling plans to meet these specific needs. To overcome this disadvantage, closed-form solutions to determine the AOQL- and LTPD-indexed single sampling plans using Artificial Neural Networks (ANNs) are proposed. To determine the closed-form solutions, feed-forward neural networks with sigmoid neural function are trained by back propagation algorithm. From the weight and bias values of these trained ANNs, the closed-form solutions to determine the sampling plans are obtained. Numerical examples are provided to demonstrate the proposed method. The proposed method does not involve table look-ups or complex calculations. Sampling plan can be determined by using this method, for any required AOQL, LTPD, lot size and process average. Suggestions are provided to duplicate this idea for applying to other standard sampling table schemes.