Statistical Inference for a Lognormal Step-Stress Model With Type-I Censoring

We consider a
k
-step step-stress model under Type-I censoring. We obtain the maximum likelihood estimates (MLE) of the parameters assuming a cumulative exposure model with lifetimes being lognormal distributed. A two-stage algorithm integrating a modified simulated annealing algorithm with a Newton-Raphson method is proposed to compute the MLE of the parameters. We also derive the confidence intervals for the parameters using asymptotic distributions, a likelihood ratio test, and parametric bootstrap resampling methods. The performance of the point and interval estimation of the parameters are evaluated through a Monte Carlo simulation study.