Reliability evaluation for Weibull distribution with heavily Type II censored data
Quality and Reliability Engineering International · 2024-10 · 40 (6) · 3620-3642
DOI: 10.1002/qre.3570 ↗Kurz gefasst
Fachartikel in der Zeitschrift Quality and Reliability Engineering International, online erschienen am 06.05.2024. Verfasst von Mengyu Liu und 2 weiteren Autor:innen. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 41 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
The lifetime data collected from the field are usually heavily censored, in which case, getting an accurate reliability evaluation based on heavily censored data is challenging. For heavily Type‐II censored data, the parameters estimation bias of traditional methods (i.e., maximum likelihood estimation (MLE) and least squares estimation (LSE)) are still large, and Bayesian methods are hard to specify the priors in practice. Therefore, considering the existing range of shape parameter for Weibull distribution, this study proposes two novel parameter estimation methods, the three‐step MLE method and the hybrid estimation method. For the three‐step MLE method, the initial estimates of shape and scale parameters are first respectively derived using MLE, then are updated by the single parameter MLE method with the range constraint of shape parameter. For the hybrid estimation method, the shape parameter is estimated by the LSE method with the existing range constraint of shape parameter, then the scale parameter estimate can be obtained by MLE. On this basis, two numerical examples are performed to demonstrate the consistency and effectiveness of the proposed methods. Finally, a case study on turbine engines is given to verify the effectiveness and applicability of the proposed methods.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Liu, M., Zheng, H., & Yang, J. (2024). Reliability evaluation for Weibull distribution with heavily Type II censored data. Quality and Reliability Engineering International, 40(6), 3620-3642. https://doi.org/10.1002/qre.3570
Vancouver
Liu M, Zheng H, Yang J. Reliability evaluation for Weibull distribution with heavily Type II censored data. Quality and Reliability Engineering International. 2024;40(6):3620-3642. doi:10.1002/qre.3570
BibTeX
@article{Liu2024,
author = {Liu, Mengyu and Zheng, Huiling and Yang, Jun},
title = {Reliability evaluation for Weibull distribution with heavily Type II censored data},
journal = {Quality and Reliability Engineering International},
year = {2024},
volume = {40},
number = {6},
pages = {3620--3642},
doi = {10.1002/qre.3570}
}
EUVIMED