Analysis of MTBF evaluation methods for small sample sizes
Hongsheng Pan, Anwei Sheng, Zhuojian Wang, Xinmin Han
Abstract
Hongsheng Pan, Anwei Sheng, Zhuojian Wang, Xinmin Han
Abstract
The accurate estimation of Mean Time Between Failures (MTBF) may be necessary under small sample size conditions and can be difficult. This article proposes combinations of two methods to calculate the cumulative probability using hierarchical Bayesian estimation and the mean rank order method, and two approaches to estimate parameters of the Weibull distribution using the ε-support vector regression (ε-SVR) and least square method. The relative error of these methods against a simulation are used to measure the accuracy of the parameter estimation where, according to the definition of MTBF, the expectation of the Weibull distribution is taken as the estimated value of MTBF. The article suggests approaches for further research in the reliability of aviation equipment.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The accurate estimation of Mean Time Between Failures (MTBF) may be necessary under small sample size conditions and can be difficult. This article proposes combinations of two methods to calculate the cumulative probability using hierarchical Bayesian estimation and the mean rank order method, and two approaches to estimate parameters of the Weibull distribution using the ε-support vector regression (ε-SVR) and least square method. The relative error of these methods against a simulation are used to measure the accuracy of the parameter estimation where, according to the definition of MTBF, the expectation of the Weibull distribution is taken as the estimated value of MTBF. The article suggests approaches for further research in the reliability of aviation equipment.
Key concepts: Mean time between failures, Weibull distribution, Statistics, Reliability (semiconductor), Mean squared error, Sample size determination, Bayesian probability, Computer science