Modal Parameter Identification And Noise Reduction Of Flood Discharge Structure Vibration Response Based On Modern Time-frequency Method

It is an issue that energy dissipation design and safe operation of high head and high flow arch dam in recent years. The real modal parameters of dam reflect an important indicator of health status. First, the traditional modal test requires a larger prototype artificial excitation, not only time consuming, but also may cause damage to the structure. The operational modal parameter of hydraulic structure with the output response only under flood discharge excitation is still a new topic. The complex work environment and a strong background noise interference on the extraction of modal parameters of great impact. So, the paper is committed to research the signal de-noising and vibration modal parameter identification. The principal achievements of this dissertation are as follows:(1) The operational modal parameter of hydraulic structure with the output response only under flood discharge excitation is identified based on EMD and ARMA method. First, based on the statistic of white noise using the Empirical Mode Decomposition, an adaptive de-noising algorithm that the false component of Intrinsic Mode Function can be judged by IMF is proposed. Modal parameters are identified with Empirical Mode Decomposition and Auto Regressive Moving Average method. The numerical experiment demonstrates that the method is effective. Last, based on the prototype observational data of an arch dam on the condition of flood discharge, the operational modal parameters of the arch dam are identified. The identified results are useful. The methods are simple, effective and of high precision, it is specially fit for modes parameters of a powerhouse structure.(2) A method to identify modal parameter for discharge structure is presented using EMD,wavelet de-noising and Hilbert Transform method, according to measurement data of the response time histories under discharge excitation. Firstly, the signal with noise of the arch dam is decomposed into a series of intrinsic mode function. And wavelet de-noising technology is used to reduce noise on the intrinsic mode function. Secondly, the modal parameter is identified by Hilbert Transform. The modal parameter is identified based on the hydroelastic Laxiwa arch dam. Finally, the method is used to identify frequency and damping of Ertan arch dam. The simulation results demonstrate that the method given in the paper can effectively improve the identification precision of modal frequency and damping ratio of discharge structure. That is a important reference value for the high arch dam real-time monitoring.