Traditional multiple deferred state sampling plan by attribute is a plan with crisp parameter. In this plan, both the current lot information and the future successive lots information are utilised on sentencing the current lot. In the present paper, firstly the authors propose a fuzzy multiple deferred state (FMDS) sampling plan by attribute when the fraction of defective items (p) is a fuzzy number. Then the proposed plan is developed to the state where the inspection is not perfect. Since p is assumed to be imprecise, the fuzzy probability theory is used to plot the operating characteristic (OC) curve of the proposed plan. It was seen that the OC curve of the proposed plan has upper and lower bounds (say FOC band). Some practical examples are given to make more realization. Finally, it was made a comparison between FMDS(0,1,2)-plans with and without the inspection errors (type 1 and 2 errors). The obtained results showed that the inspection errors have a negative influence on the accuracy and performance of the FOC band of the proposed plan.

Keywords