Abstract
Reliability assessment for hard disk drives (HDDs) is important yet difficult for manufacturers. Motivated by the fact that the particle accumulation in the HDDs, which accounts for most HDD catastrophic failures, is contributed from the internal and external sources, a counting process with two arrival sources is proposed to model the particle cumulative process in HDDs. This model successfully explains the collapse of traditional ALT approaches for accelerated life test data. Parameter estimation and hypothesis tests for the model are developed and illustrated with real data from a HDD test. A simulation study is conducted to examine the accuracy of large sample normal approximations that are used to test existence of the internal and external sources. © 2012 Elsevier Ltd. All rights reserved.
| Original language | English |
|---|---|
| Pages (from-to) | 110-118 |
| Journal | Reliability Engineering and System Safety |
| Volume | 109 |
| DOIs | |
| Publication status | Published - Jan 2013 |
Research Keywords
- Accelerated life test (ALT)
- Hard disk drive (HDD)
- Nelson-Aalen plot
- Non-homogeneous poisson process (NHPP)
- Poisson distribution
Fingerprint
Dive into the research topics of 'Reliability evaluation of hard disk drive failures based on counting processes'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver