STATISTICAL MODELING OF DEVICE UPTIME UNDER INCOMPLETE SAMPLE CONDITIONS

Authors

  • Shakhodat Sadriddinovna Murodova

    Tashkent Chemical-Technological Institute image/svg+xml

Keywords: continuous operation time, incomplete sample, maximum likelihood estimation, Veibull distribution, censoring, Fisher information, bias correction

Abstract

The paper considers the issues of statistical modeling of device continuous operation time under incomplete (right-censored) sample conditions. It is proved that under small sample sizes and high censoring rates, maximum likelihood estimates for the Weibull shape parameter exhibit significant positive bias. Leveraging Fisher information and the number of observed failures, an analytical correction factor (Kcorr) and refined confidence intervals were derived. The proposed method reduces estimation bias by up to 20 times, significantly improving reliability assessment accuracy.

References

1. Lawless J.F. Statistical models and methods for lifetime data. 2nd ed. – Hoboken: John Wiley & Sons, 2003.

2. Nelson W. Accelerated testing: statistical models, test plans, and data analysis. – Hoboken: John Wiley & Sons, 2004.

3. Meeker W.Q., Escobar L. A., Pascual F. G. Statistical methods for reliability data. 2nd ed. – Hoboken: John Wiley & Sons, 2022.

4. Чимитова Е.В. Математическое и алгоритмическое обеспечение статистического анализа данных типа времени жизни: дис. ... докт. техн. наук: 05.13.17. — Новосибирск: НГТУ, 2016.

5. Лемешко Б.Ю. Критерии проверки статистических гипотез: руководство по применению. – Москва: ИНФРА-М, 2021.

6. Kalbfleisch J. D., Prentice R. L. The statistical analysis of failure time data. 2nd ed. – Hoboken: John Wiley & Sons, 2002.

7. Gentle J.E. Random number generation and Monte Carlo methods. 2nd ed. – New York: Springer, 2003.

8. Rao C.R. Linear statistical inference and its applications. 2nd ed. – New York: John Wiley & Sons, 2002.