An integrated approach to estimate storage reliability with initial failures based on E-Bayesian estimates
2017 (English)In: Reliability Engineering & System Safety, ISSN 0951-8320, E-ISSN 1879-0836, Vol. 159, 24-36 p.Article in journal (Refereed) Published
Storage reliability that measures the ability of products in a dormant state to keep their required functions is studied in this paper. For certain types of products, Storage reliability may not always be 100% at the beginning of storage, unlike the operational reliability, which exist possible initial failures that are normally neglected in the models of storage reliability. In this paper, a new integrated technique, the non-parametric measure based on the E-Bayesian estimates of current failure probabilities is combined with the parametric measure based on the exponential reliability function, is proposed to estimate and predict the storage reliability of products with possible initial failures, where the non-parametric method is used to estimate the number of failed products and the reliability at each testing time, and the parameter method is used to estimate the initial reliability and the failure rate of storage product. The proposed method has taken into consideration that, the reliability test data of storage products containing the unexamined before and during the storage process, is available for providing more accurate estimates of both the initial failure probability and the storage failure probability. When storage reliability prediction that is the main concern in this field should be made, the non-parametric estimates of failure numbers can be used into the parametric models for the failure process in storage. In the case of exponential models, the assessment and prediction method for storage reliability is presented in this paper. Finally, a numerical example is given to illustrate the method. Furthermore, a detailed comparison between the proposed and traditional method, for examining the rationality of assessment and prediction on the storage reliability, is investigated. The results should be useful for planning a storage environment, decision-making concerning the maximum length of storage, and identifying the production quality.
Place, publisher, year, edition, pages
2017. Vol. 159, 24-36 p.
Bayesian networks; Behavioral research; Decision making; Digital storage; Failure analysis; Forecasting; Maximum likelihood; Maximum likelihood estimation; Numerical methods; Probability; Program processors; Reliability; Statistical tests, Bayesian estimations; Integrated techniques; Least squares estimation; Non-parametric estimates; Nonparametric methods; Operational reliability; Reliability functions; Storage reliability, Parameter estimation
IdentifiersURN: urn:nbn:se:hig:diva-22874DOI: 10.1016/j.ress.2016.10.024ISI: 000392897600003ScopusID: 2-s2.0-84994571996OAI: oai:DiVA.org:hig-22874DiVA: diva2:1050291
Funding Agency: National Natural Science Foundation of China Grant Number: 70571018 and Grant Number: 11401007; Natural Science Foundation of Jiangsu Province of China Grant Number: BK20150455 2016-11-282016-11-282017-03-13Bibliographically approved