Multiattribute acceptance sampling inspection plan

Abstract

This thesis is concerned with an attribute acceptance sampling for multicharacteristic items utilizing the ratio $ρ_i$``s between adjacent numbers of occurrences of characteristic defects and gives the desired protection at LTPD quality of each characteristic. The use of the ratio $ρ_i$``s instead of pareto relationship $\bar{ρ}$, the average of $ρ_i$``s, is recommended after examining several sampling plans. Sampling plans having acceptance number one and chain sampling type inspection plans are employed in order to improve the poor shape of OC curves of minimum size sampling plans. Then a simple modified multiattribute acceptance sampling inspection procedure is proposed. Finally the minimum sample size plan for given AOQL is considered after extending the concept of $ρ_i$``s to the processes operating in a random manner.