|
| << vorige | volgende >> |
Tijdschrift beschrijving
Alle jaargangen van het bijbehorende tijdschrift
Alle afleveringen van het bijbehorende jaargang
Alle artikelen van de bijbehorende aflevering
Details van artikel 7 van 23 gevonden artikelen
| Titel: |
Gamma distribution parameter estimation for field reliability data with missing failure times |
| Auteur: |
Coit, David W. Jin, Tongdan |
| Verschenen in: |
IIE transactions |
| Paginering: | Jaargang 32 (2000) nr. 12 pagina's 1161-1166 |
| Jaar: |
2000-12-01 |
| Inhoud: |
Maximum likelihood estimators have been developed for the gamma distribution when there is missing time-to-failure information. Data sets with missing time-to-failure data can arise from field data collection systems that rely on recorded observations of the system by the operators and maintenance personnel. In many regards, this type of data is highly desirable because it implicitly accounts for all actual usage and environmental stresses. Unfortunately the component times-to-failure are not always recorded for fielded systems because of a lack of elapsed time meters, unsatisfactory data reporting requirements, or incomplete or lost information. When only data of this type is available, it creates a non-standard form of da'ta censoring and it has generally not been possible to fit most common time-to-failure distributions. Reliability practitioners have sometimes made unsubstantiated simplifying assumptions so the data can be used. In this paper, a more rigorous approach is presented. Maximum likelihood estimators are derived and demonstrated for the gamma distribution based on merged data records where the individual failure times have not been recorded. These results are important because the gamma distribution can model diverse time-to-failure behavior. This provides a particularly useful tool for data sets that may otherwise not be satisfactorily analyzed. |
| Uitgever: |
Taylor & Francis |
| Bronbestand: |
Elektronische Wetenschappelijke Tijdschriften |
Details van artikel 7 van 23 gevonden artikelen
| << vorige | volgende >> |