摘要
•An innovative sampling system with adaptable rule-switching mechanism is created.•The proposed systems OC function is derived using Markov chain transition matrix.•Nonlinear optimization models are constructed to determine optimal system designs.•Several analyses on performance metrics are investigated, compared and discussed.•An interactive cloud-computing application is created to enhance the applicability.
The Quick-Switch Sampling System (QSS) has gained significant attention for its integration of both a normal Single Sampling Plan (SSP) and a tightened SSP to handle lot dispositions. Nevertheless, existing QSS implementations are limited by rigid rule-switching mechanisms. In practical scenarios, paticularly when a delivered lot is rejected due to potential process degradation, practitioners often need to apply the tightened SSP multiple times to ensure consistent quality. To address this limitation, we propose a generalized and flexible framework based on the original QSS, referred to as QSS-r, designed to meet modern quality control needs. The parameter r, a specifiable positive integer, allows practitioners to define the required number of consecutive acceptances under the tightened SSP before transitioning back to the normal SSP. We derive the operating characteristic and average sample number functions for QSS-r, leveraging the process yield index Spk as the foundation. Analytical insights form the basis for nonlinear optimization models used to identify optimal system designs. A comprehensive evaluation of various QSS-r configurations reveals their distinct strengths and weaknesses. Finally, we demonstrate the real-world applicability and feasibility of QSS-r through a case study involving silicon-based solar cells, highlighting its potential to enhance quality assurance in industrial applications.