Improved performance of damage localization and severity estimation via combining multiple damage algorithms /
To date, many non-destructive damage detection algorithms have been proposed and developed on various theoretical and experimental grounds. So far, however, it is observed that each and every algorithm has its limitations as well as advantages when applied to practical problems. To resolve these lim...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1999.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=733670231&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | To date, many non-destructive damage detection algorithms have been proposed and developed on various theoretical and experimental grounds. So far, however, it is observed that each and every algorithm has its limitations as well as advantages when applied to practical problems. To resolve these limitations, two ways to improve the performance of damage localization and severity estimation are explored in this study: (a) by enhancing the individual algorithms via utilizing more sensitive response measures to damage and (b) by combining multiple damage predictions from different algorithms. To these ends, specific algorithms using various types of response measures are developed from the generic expression of the two prominent damage detection algorithms: namely the algorithms that utilize the changes in the strain energy distribution and the flexibility distribution. The performance of damage detection using the specific algorithms is evaluated using data generated from numerical simulations. Then, various logical rules for combining the algorithms are proposed and the performances of damage detection using these rules are compared with the performance of individual algorithms. It is demonstrated that some response measures produce more accurate and consistent damage predictions than commonly used modal response measures and the damage prediction of the consensus of multiple damage algorithms may improve the performance of damage localization and severity estimation. |
|---|---|
| Item Description: | Vita. "Major Subject: Civil Engineering". |
| Physical Description: | xvii, 237 leaves : illustrations ; 28 cm. Issued also on microfiche from University Microfilm Inc. |
| Bibliography: | Includes bibliographical references (leaves 225-233). |