Abstract
Section 1 of the full paper briefs the probability density function, which can be estimated by the existing methods in Refs. 4 and 5 and by the two new methods; in section 1, eq. (2), which is related to augmented reliability, is particularly important. Subsection 2.1 establishes the new adaptive kernel density approximation method; in subsection 2.1, eq. (8) can be used for estimating the probability density function. Subsection 2.2 establishes the new polynomial approximation method; in subsection 2.2, eq. (14) can be used for estimating the probability density function. Section 3 gives the simulation results of three examples and their analysis. The simulation results are presented in: Figs. 1 and 2 for example 1; Figs.3 and 4 for example 2. The analysis of the results shows mainly and preliminarily that the existing maximum entropy method and the new polynomial approximation method for analyzing global reliability sensitivity are more robust than the other two methods.
| Original language | English |
|---|---|
| Pages (from-to) | 664-668 |
| Number of pages | 5 |
| Journal | Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University |
| Volume | 27 |
| Issue number | 5 |
| State | Published - Oct 2009 |
Keywords
- Adaptive kernel density approximation
- Maximum entropy
- Polynomial approximation
- Probability density function
- Reliability
- Sensitivity analysis