The Health Monitoring Method of Concrete Dams Based on Ambient Vibration Testing and Kernel Principle Analysis

Abstract

The ambient vibration testing (AVT) measurement of concrete dams on full-scale can show the practical dynamic properties of structure in the operation state. For most current researches, the AVT data is generally analyzed to identify the structural vibration characteristics, that is, modal parameters. The identified modal parameters, which can provide the global damage information or the damage location information of structure, can be used as the basis of structure health monitoring. Therefore, in this paper, the health monitoring method of concrete dams based on the AVT is studied. The kernel principle analysis (KPCA) based method is adopted to eliminate the effect of environmental variables and monitor the health of dam under varying environments. By taking full advantage of the AVT data obtained from vibration observation system of dam, the identification capabilities and the warning capabilities of structural damage can be improved. With the simulated AVT data of the numerical model of a concrete gravity dam and the measured AVT data of a practical engineering, the performance of the dam health monitoring method proposed in this paper is verified.