초록 열기/닫기 버튼
Purpose: This paper reports on the corresponding guidelines to help users select a suitable estimation method. The guidelines were derived by evaluating the reliability estimation performance of different estimation methods. Methods: Weibull distribution parameters were calculated using the least squares and maximum likelihood estimators, aling with the Bayesian methods. The scale and shape parameters were estimated to calculate the life-time. Finally, the analysis of variance was performed to compare the accuracy of the various methods. Results: Bayesian methods, which employed prior information, exhibited a relatively high performance for all sample sizes. As the sample size increased, the performance was similar to that of the least squares and maximum likelihood estimators. The performance of the Bayesian methods fluctuated according to the prior information. Conclusion: The reliability of various methods to analyze Weibull-distribution-based censoring data was analyzed. The results can be used in reliability assessment, to achieve the target reliability in the product development phase.
