Abstract: This research introduced a modified acceptance sampling plan to improve quality control in continuous production systems, particularly under accelerated life testing (ALT) conditions modelled by the Weibull distribution. The proposed sampling plan, known as the multiple dependent state-triple sampling plan (MDS-TSP), integrated the principles of the multiple dependent state sampling plan and triple sampling to create a more efficient and flexible inspection framework. A genetic algorithm was employed to identify the optimal plan parameters for the MDS-TSP. The performance of MDS-TSP was evaluated using the average sample number, while the impact of model parameters on the MDS-TSP solution was examined by sensitivity analysis using an orthogonal experimental design and multiple linear regression. A comparative performance analysis was conducted between the MDS-TSP and existing sampling plans using identical parameter settings. The results demonstrated that the MDS-TSP achieved a significant reduction in average sample number compared to existing sampling plans, confirming its superior efficiency. The applicability of the MDS-TSP was validated through a real dataset involving electronic device reliability. These case studies illustrate the practicality and usefulness of the proposed approach in industrial settings. Our findings support the conclusion that the MDS-TSP is a flexible, effective, and resource-efficient solution for quality control in continuous production systems under Weibull-distributed accelerated life testing conditions.

Keywords: acceptance sampling plan, average sample number, accelerated life testing, multiple dependent state sampling plan, triple sampling, Weibull distribution

Pages: 376-395