Reliability Inference in Constant Stress Accelerated Life Tests Based on the Exponentiated Half-Logistic Model

Accelerated life testing is an efficient methodology that enables the acquisition of failure time data in a shorter period by subjecting test units to different stress levels, compared to tests conducted under normal operating conditions. In this study, it is assumed that the lifetime of a product under a constant stress level follows an exponentiated half-logistic distribution and the model parameters, reliability function and mean time to failure function are estimated using eleven different frequentist methods, namely, maximum likelihood, least squares, weighted least squares, Cramér-von Mises, Anderson-Darling, right-tail Anderson-Darling, left-tail Anderson-Darling, second-order left-tail Anderson-Darling, maximum product of spacings, minimum spacing absolute distance, and minimum spacing absolute-log distance. The performances of the estimators are compared through an extensive Monte Carlo simulation study. Furthermore, a real dataset is analyzed to evaluate the practical effectiveness of the proposed methods.