الفهرس 
introductionOnly 14 pages are availabe for public view
 |  | from | 241 |  |  |
| | | from | 241 | | |
Abstract
In acceptance sampling plans, both the producer and consumer aim at making sure that the submitted lots are good (i.e. they meet all quality criteria agreed upon between the producer and consumer). There are two ways to do so:
Total inspection: that’s examining each unit in the lot, removing the non-conforming items and replacing them with good quality units.
Drawing a random sample or more from the lot, examining its units and taking a decision whether to accept or reject the lot.
Depending on the number of the drawn samples, the types of plans for this study are: single, double, multiple and group acceptance sampling plans. According to this, this research aims at studying acceptance sampling plans when the lifetime of the product is a random variable that has a probability distribution: generalized inverse Weibull distribution introduced by (de Gusmão et al. (2009)), and (Al-Omari (2016)) in single sampling plan only. This study aims at introducing this distribution in different acceptance sampling plans at various values of the parameters. The research is also concerned with developing two distributions: the first one is related to generalize inverse Weibull. This distribution was suggested by using Azzalini’s method (1985) to create a new shape parameter. The second distribution, compound inverse Rayleigh distribution is related to inverse Rayleigh. This distribution is suggested by compounding the gamma(α,β) with inverse Rayleigh (θ). The main objective of this research is to achieve the minimum sample sizes required for a given acceptance limit when the consumer’s risk and the termination time ratio are specified. The operating characteristic values based on different values of quality level are obtained and minimum ratios of the actual median life to specified median life at the specified producer’s risk (0.05) are determined. Some plans are compared: single with group plans and multiple with single and double. Some comparisons are also made between the results of these distributions; specifically the results obtained from skew-generalized inverse Weibull and generalized inverse Weibull as well as the results of compound inverse Rayleigh with inverse Rayleigh in single sampling plans.
The results showed that the performance of skew-generalized inverse Weibull in sampling plans for accepting the lot is better than the performance of generalized inverse Weibull. In addition, the performances of multiple sampling plans for accepting the lot is better than the performance of single and double sampling plans.