• 품질경영학회지
• /
• 제21권1호
• /
• pp.108-120
• /
• 1993

In this article an economic life test sampling plan is considered for repairable products when the products in each lot have the same interfailure time distribution, but the mean time between failure (MTBF) of a lot varies from lot to lot according to a known prior distribution. A cost model is constructed which consists of test cost, accept cost, and reject cost. Determination of the optimal plan which minimizes the expected average cost per lot is discussed. Numerical examples are presented to illustrate the use of the proposed sampling plans and sensitivity analyses for parameters of the prior distribution are performed.

• Journal of the Korean Data and Information Science Society
• /
• 제5권2호
• /
• pp.75-85
• /
• 1994

계단 충격 가속수명시험에서 통계적 추론을 위해 가정하는 수명분포에 대한 적합도검정을 Kolmogorov-Smirnov, Kuiper, Watson, Cramer-von Mises, Anderson-Darling과 같은 비모수적 검정통계량에 대하여 몬테칼로 방법을 이용한 기각치를 구하고, 검정력 측면에서 비교, 연구한다.

• 품질경영학회지
• /
• 제29권3호
• /
• pp.49-65
• /
• 2001

This paper presents new compromise ALT plan which is applied to situations that true relationship between stress and parameters is not known exactly. The assumed failure distribution of this study is one of location-scale family, i. e., exponential, Weibull, and lognormal distributions which have been ones of the popular choices of failure distributions. The method of applying the stress is constant, and the censoring mechanism is Type I censoring. Compared with existing compromise plans under true simple linear model in terms of statistical efficiency, the efficiency of new compromise plan is better than the corresponding other compromise ones in most cases. For case when true model is quadratic, this plan can be used without any severe loss in statistical efficiency. The proposed new compromise ALT plan is illustrated with a numerical example and sensitivity analyses are conducted to study effects of pre-estimates of design parameters.

• Journal of the Korean Data and Information Science Society
• /
• 제13권2호
• /
• pp.235-242
• /
• 2002

In this paper, we develop noninformative priors that are used for estimating the ratio of failure rates under Freund's bivariate exponential distribution. A class of priors is found by matching the coverage probabilities of one-sided Baysian credible interval with the corresponding frequentist coverage probabilities. Also the propriety of posterior under the noninformative priors is proved and the frequentist coverage probabilities are investigated for small samples via simulation study.

• Journal of the Korean Data and Information Science Society
• /
• 제13권2호
• /
• pp.243-250
• /
• 2002

This paper considers the problem of estimating parameters of the bivariate exponential distribution with a location parameter for a two-component shared parallel system using component data from system-level life test terminated at the time of the prespecified number of system failure. In the system-level life testing, there are three patterns of failure types ; 1) both component failed 2) both component censored 3) one is failed and the other is censored. In the third case, we assume that the failure time might be known or unknown. The maximum likelihood estimators are obtained for the case of known/unknown failure time when the other component is censored.

• 품질경영학회지
• /
• 제20권1호
• /
• pp.22-32
• /
• 1992

This thesis compares life test acceptance sampling plans under lifetime has an exponential distribution. Various practical considerations may lead a user adopt a two-stage, or double sampling, test procedure. Hewett and Spurrier(1983) provided a survey of two-stage methods, as well as examples of experiments for which a two-stage procedure would be appropriate. The plans are compared in terms of the expected number of failures, and the expected time required to reach a dicision. Computational experiments are conducted and the results are tabulated to provide guidelines for selecting an appropriate plan for a given situation.

• 품질경영학회지
• /
• 제23권4호
• /
• pp.13-27
• /
• 1995

This paper considers the design of life test sampling plans based on Type II censored accelerated life tests. The lifetime distribution of products is assumed to be exponential. An estimator of acceleration factor between use condition stress and an accelerated level of stress higher than use condition stress is assumed to be known. The critical value for lot acceptance and the number of failure before censoring which satisfy the producer's and consumer's risk requirements are determined. The properties of the proposed life test sampling plans are investigated.

• Journal of the Korean Statistical Society
• /
• 제23권2호
• /
• pp.421-444
• /
• 1994

Reliability acceptance sampling is concerned with whether to accept or reject a collection of items based upon the information obtained from life testing. Although various reliability acceptance sampling plans have been developed, little is known about their relatvie performances. This paper compares reliability acceptance sampling plans under Type II censoring, Hybrid censoring, and Time-Truncated Type II censoring assuming that the lifetimes of items in a lot follow an exponential distribution. The three plans are compared in terms of the power, the expected number of failures, and the expected time required to reach a decision. Computational experiments are conducted and the results are tabulated to provide guidelines for selecting an appropriate plan for a given situation.

• 응용통계연구
• /
• 제11권1호
• /
• pp.185-192
• /
• 1998

Freund(1961)가 제안한 이변량지수분포는 두 부품으로 이루어진 병렬체계의 상호종속적인 부품의 수명을 해석하는 등에 응용될 수 있어 널리 이용되고 있다. 본 연구에서는 위치모수를 가지는 이변량지수분포를 Freund 모형을 일반화시키는 차원에서 제안하고 모형의 통계적 성질 및 모수에 대한 최우추정량을 구하였다. 또한 최우추정량을 수정하여 편의는 감소시킬 수 있는 새로운 추정량을 제안하였다.

• Communications for Statistical Applications and Methods
• /
• 제3권3호
• /
• pp.291-297
• /
• 1996

In this paper, we propose the jackknife estimator and the bootstrap estimator of Gini index of the two-parameter exponential distribution when the location parameter $\theta$ is unknown and the scale parameter $\sigma$is known. Sinilarly, we propose the bias location parameter $\theta$ and the scale parameter $\sigma$ are unknown. The bootstrap estimator is more efficient than the other estimators when the location parameter $\theta$is unknown and the scale parameter $\sigma$ is known, and the bias corrected estimator is more efficient than the MLE when both the location parameter $\theta$ and the scale parameter $\sigma$are unknown.