| China is one of the countries with the most serious earthquake disasters in the Pacific rim and the Mediterranean Sea, Himalaya’s two major seismic belts. At the same time, China is also one of the countries that have built and built concrete gravity dams in the world. For the seismic analysis of concrete gravity dam in service, especially the concrete gravity dam, which has been tested by the earthquake, the selection of the dynamic parameters of the concrete gravity dam is a problem worthy of further study. Based on the strong earthquake observation of concrete gravity dam, the modal parameter identification of the dam structure is carried out. Concrete gravity dam is a complicated multiple degree of freedom spatial structure, the traditional matrix model and parameter correction model correction method for the finite element model of concrete gravity dam correction needs a large number of finite element calculation, the calculation efficiency is low, not strong applicability of the concrete gravity dam of large engineering structure etc.. In this paper,considering the characteristics of concrete gravity structure itself and the strong earthquake observation data,combined with the theory of modal analysis and intelligent optimization algorithm for the dynamic finite element model is modified with the concrete gravity dam under strong earthquake observation data, the main research contents and results are as follows:(1) Some of the common seismic records denoising methods are introduced, meanwhile, proposes the denoising method of strong motion observation data of concrete gravity dam based on PCA and wavelet filtering, and wavelet denoising method and the method of soft and hard threshold denoising effect were compared by simulation; with strong motion data after denoising based on the proposed automatic domain decomposition based on (AutoFrequencyDomainDecomposition, AFDD) parameter identification method of concrete gravity dam modal method, and adopts the method of strong motion observation records of a concrete gravity dam measured modal parameters.(2) The reservoir dam - based 3D dynamic finite element analysis relates to the water dam and foundation dam dynamic interaction and viscous elastic artificial boundary theory are introduced, and on this basis, the research of sensitivity analysis methods for finite element model updating parameter selection and parameters of concrete. The above analysis process is realized by a practical engineering example.(3) The modal correlation as the criterion, the concrete gravity dam dynamic finite element model updating problem is transformed into the problem of minimizing the difference between a strong modal parameter identification and calculation of finite element modal parameters of the. Considering the characteristics of concrete gravity structure itself and the strong earthquake observation data, the experimental design ability of Latin hypercube sampling, multi output support vector machine (Multi-output Support Vector Machine, M-S VM) fitting ability of nonlinear implicit relations between multivariate algorithm and genetic algorithm (Genetic Algorithm GA) neural network search ability to overflow a strong earthquake observation and M-SVM concrete gravity dam dynamic finite element model updating method based on. The finite element model of a concrete gravity dam is modified by this method, and the calculation efficiency is improved effectively. |