23º SINAPE - Simpósio Nacional de Probabilidade e Estatística

Dados do Trabalho


Título

BAYESIAN INFERENCE FOR THE STEP-STRESS ACCELERATED TEST UNDER WEIBULL AND GAMMA DISTRIBUTIONS

Resumo

Step-stress procedure is a quite popular accelerated test used for analyzing lifetime of highly reliable components. This paper presents simple and multiple step-stress accelerated life tests for the cumulative exposure model using complete and censored data. Assuming that the lifetimes of test item follow the Weibull and Gamma distributions, the maximum likelihood estimation and Bayesian approaches are used to estimate the parameters. Some proposed priors are investigated and compared under symmetric and asymmetric loss functions such as squared, linex and general entropy loss functions. It is also considered Type II censoring scheme. A sensitivity analysis of these Bayesian and maximum likelihood estimators are also presented by simulation Monte Carlo study, and a real data set is analyzed for illustrative purposes.

Palavras-chave

Bayesian, accelerated test, step-stess,Weibull, Gamma, cumulative exposure, maximal entropy prior, loss function, general entropy, Linex, MCMC.

Área

Análise de Sobrevivência

Autores

Fernando Antonio Moala, Karlla Delalibera Chagas, Narayanaswamy Balakrishnan