DESIGN AND OPTIMIZATION OF A FUZZY CHAIN SAMPLING PLAN FOR TEXTILE QUALITY CONTROL

Abstract

Quality control is essential in the textile industry, as defects in fabrics, yarns, dyeing, and finishing directly affect production efficiency and customer satisfaction. This study proposes a fuzzy chain sampling plan to handle the variability commonly found in textile processes. A case study has been carried over the data obtained from original  inspection reports by applying fuzzy chain sampling techniques for  varying: sample sizes, number of defectives, acceptance number and   chain lengths. The defect proportion is modeled as a triangular fuzzy number, and α-cut decomposition is used to generate fuzzy acceptance intervals.   By using single-sampling plan  acceptance probabilities are computed through the Poisson distribution for fuzzy acceptance intervals. A decision rule is chosen by selecting an appropriate defuzzification or fuzzy decision method. One option is to require that the centroid  value of the chain acceptance probability exceeds a chosen threshold. Chain acceptance is determined based on consecutive acceptable samples. The plan is optimized by adjusting the chain length and acceptance number to balance producer’s and consumer’s risks while minimizing inspection effort. The highest probability of acceptance when the defect proportion lies roughly between 5% and 8% is also determined. The resulting plan offers a more flexible and reliable approach to quality assessment in textile manufacturing.