|
| A novel sampling method with error estiomation for the reduced basis method |
| Received:May 08, 2009 Revised:October 28, 2010 |
| View Full Text View/Add Comment Download reader |
| DOI：10.7511/jslx201105025 |
| KeyWord:reduced basis method dynamics sampling method eigenvalue error transient response |
| Author | Institution |
| 张正 |
湖南大学 汽车车身先进设计与制造国家重点实验室,长沙 |
| 韩旭 |
湖南大学 汽车车身先进设计与制造国家重点实验室,长沙 |
| 姜潮 |
湖南大学 汽车车身先进设计与制造国家重点实验室,长沙 |
|
| Hits: 1940 |
| Download times: 1433 |
| Abstract: |
| The vector-space approximation sampling method is proposed here for the reduced basis method in which the precision of solution is influenced by completeness of basis space. The method can enable the corresponding eigenvector obtained in every step to have the biggest angle with the space constructed by early sampling eigenvectors, and then it makes the constructed eigenvector basis space have the weakest correlation and thus be more complete. Furthermore, this method can bring on a prior eigenvalue error bound which can ensure the eigenvalue errors produced by the reduced basis methods to be limited in the previously set scopes. The numerical example demonstrates the validity of the proposed sampling method. |