Estimation of time-to-failure distribution derived from a degradation model using fuzzy clustering

✍ Scribed by Shuo-Jye Wu; Tzong-Ru Tsai

• Publisher: John Wiley and Sons
• Year: 2000
• Tongue: English
• Weight: 100 KB
• Volume: 16
• Category: Article
• ISSN: 0748-8017
• DOI: 10.1002/1099-1638(200007/08)16:4<261::aid-qre333>3.0.co;2-3

✦ Synopsis

Some life tests are terminated with few or no failures. In such cases, a recent approach is to obtain degradation measurements of product performance that may contain some useful information about product reliability. Generally degradation paths of products are modeled by a nonlinear regression model with random coefficients. If we can obtain the estimates of parameters under the model, then the time-to-failure distribution can be estimated. In some cases, the patterns of a few degradation paths are different from those of most degradation paths in a test. Therefore, this study develops a weighted method based on fuzzy clustering procedure to robust estimation of the underlying parameters and time-to-failure distribution. The method will be studied on a real data set.