TY - JOUR
TI - Statistical inference for a Copula-based simple step-stress accelerated life model
AU - Wang Ying
AU - Guo Yongfei
JN - Thermal Science
PY - 2026
VL - 30
IS - 2
SP - 951
EP - 960
PT - Article
AB - In this paper, we undertake a statistical analysis of a simple step-stress  accelerated competing failure model. The prevailing hypothesis posits that  the competing failure modes are mutually independent. However, empirical  evidence suggests that failures tend to be dependent in practice.  Consequently, we hypothesize that the failure mechanisms are interconnected.  The dependence structure examined in this article is constructed by  Gumbel-Hougard Copula and Clayton Copula, and different construction  methods are compared. The estimation of unknown parameters is achieved  through the maximum likelihood estimation method. The precision of these  estimates is subsequently evaluated through Monte Carlo simulations. The  simulation results demonstrate that: The Copula theory plays a monumental  role in the study of the correlation of competitive failure mechanisms. In  this regard, the Gumbel-Hougard Copula demonstrates superior performance in  comparison to the Bivariate Clayton Copula within the same framework.  Finally, the methods of inference discussed are illustrated with a real  dataset.
DO - 10.2298/TSCI2602951W
ER -