A Novel Damage Sensitive Feature Extraction Method of the Concrete Dam
| Autor: | Xi Zhu, Cao Enhua, Tengfei Bao, Hui Li |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Computer science
Feature extraction Geotechnical Engineering and Engineering Geology computer.software_genre Domain (software engineering) Constraint (information theory) Vibration Identification (information) Feature (computer vision) Singular value decomposition Structural health monitoring Data mining computer Civil and Structural Engineering |
| Zdroj: | Iranian Journal of Science and Technology, Transactions of Civil Engineering. 46:2173-2186 |
| ISSN: | 2364-1843 2228-6160 |
| DOI: | 10.1007/s40996-021-00709-5 |
| Popis: | The damage identification of concrete dams provides a crucial alternative to assess their current conditions, and can provide continual safety evaluation throughout their whole lifespan. A novel data mining approach for damage identification which named the multi-resolution singular value decomposition-based permutation entropy (MSVD-PE) combined the eXtreme Gradient Boosting (XGBoost) is proposed to identify the different structural conditions of concrete dams. In this method, the MSVD-PE algorithm is first used to extract the damage feature from the de-noised vibration signals. Secondly, the low-rank constraint modified Laplacian Score algorithm is developed to refine the damage feature. Finally, the obtained feature is fed into the XGBoost model to accomplish the damage pattern identification. The proposed method is experimentally demonstrated to be able to recognize the different damage categories in concrete dams. Hence, the method proposed in this paper can provide reference for the safety assessment of the concrete dam in the operation and maintenance stage in the domain of Structural health monitoring. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink