|
| Application of maximum entropy method in reliability computation |
| Received:October 31, 2013 Revised:December 03, 2013 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7511/jslx201501007 |
| KeyWord:maximum entropy reliability probability density function |
| Author | Institution |
| 赖雄鸣 |
华侨大学 机电及自动化学院, 厦门 |
| 王成 |
华侨大学 计算机科学与技术学院, 厦门 |
| 张勇 |
华侨大学 机电及自动化学院, 厦门 |
| 言兰 |
华侨大学 机电及自动化学院, 厦门 |
| 缑锦 |
华侨大学 计算机科学与技术学院, 厦门 |
|
| Hits: 2102 |
| Download times: 1974 |
| Abstract: |
| As for reliability computation,arbitrary independent random variables always follow with different types of probability distribution,which bring in difficulties for computing reliability.Generally,arbitrary independent random variables are always transformed into the unified form of normal variables by means of normal transformation for the convenience of reliability analysis.But normal transformation can not definitely ensure the probability distribution feature of the original probability density function (PDF) and to some degree affects the accuracy of the computed reliability results.Aiming at this problem,the paper transforms the PDF of arbitrary independent random variables into the the unified form of the maximum entropy PDF (MEPDF) by means of Maximum entropy method,which are more accurate and then exploits the corresponding the first and second order moment approach.The proposed method has been validated by practical analyses. |
|
|
|