摘要
Quick-switch sampling (QSS) systems, which utilize process capability indices (PCIs) as benchmark measures, have garnered substantial attention in recent literature due to their effectiveness in assessing lot quality and cost-effectiveness. Nevertheless, current QSS models exhibit limited adaptability in rule-switching, diminishing their practical utility. In this paper, we introduce variables generalized QSS (GQSS) systems in two distinct forms with the aim of enhancing the rule-switching mechanism inherent in conventional QSS systems. By applying Markov chain theory and mean first passage time concepts, we determine the operational characteristics, average sample number, and average run length functions for GQSS systems based on unilateral PCIs. Furthermore, we develop non-linear optimization models to facilitate optimal system design. Through comprehensive analysis and comparative evaluation, our findings demonstrate that the proposed GQSS systems excel in detecting lot quality degradation while maintaining cost-effectiveness. To validate their feasibility, we present an industrial case study. In conclusion, the proposed GQSS systems offer a more adaptable and versatile rule-switching mechanism compared to existing QSS systems, thereby significantly broadening their practical applicability.