The Open Cybernetics & Systemics Journal

2015, 9 : 729-733
Published online 2015 June 26. DOI: 10.2174/1874110X01509010729
Publisher ID: TOCSJ-9-729

An Approximate E-Bayesian Estimation of Step-stress Accelerated Life Testing with Exponential Distribution

Liping Hu and Xianbin Wu
Junior College, Zhejiang Wanli University, Ningbo, Zhejiang, 315100, China.

ABSTRACT

The data analysis problem of step-stress accelerated life testing with exponential distribution is discussed. At the step-stress accelerated life testing conditions, an approximate E-Bayesian parameter estimation of step-stress accelerated life testing with exponential distribution is given by considering the prior distributions of the hyperparameters and using Gibbs sampling method. Finally, a simulation example is given, the results show that the Gibbs sampling method is simple and the convergence is better. E-Bayesian parameter estimation is more effective than the maximum likelihood estimation.

Keywords:

E-Bayesian estimation, step-stress life testing, exponential distribution, Gibbs sampling.