Quantitative Classification And Parameterized Analysis Of Near-fault Pulse-like Ground Motions

For decades,some earthquakes have posed a tremendous threat to the life and properties of human society.Researchers,meanwhile,have acquired near-fault seismograms and thought that the large impulse contained in ground motion is the main factor causing the severe damage to modern structures.Therefore,the research on characteristics about near-fault ground motions will provide an effective basis for considering the seismic performance of structures near the fault.This paper mainly studies the quantitative identification of near fault pulse ground motions,the spectral analysis of seismic artery impulse,and the parametric analysis of pulse.The main research contents and conclusions are as follows:(1)The existing identification methods may lead to failure of pulse recognition when the shape of the seismic pulse is irregular.So we establish an energy recognition method using the intrinsic mode function(IMF)reconstruction pulse signal,based on the adaptive principle of empirical mode decomposition(EMD).1.The similarity theory by Tanimoto is used to discard the high-frequency IMF component and reorganize the rest to form a smooth reconstruction pulse signal,which is to remove the interference caused by high frequency burr.2.The peak point method(PPM)is used to determine the pulse period and the relative energy of the pulse.3.The pulse indicator based energy is established as the basis for the recognition of the pulse type ground motion.In addition,this paper also extends the multi-pulse identification and pulse directionality.The results show that the proposed method is more accurate than the other methods in the case of irregular pulse shape and multi-pulse.(2)Dynamic response spectra analysis are utilized to validate the method,after verifying the matching degree between the recognition pulse and the original ground motion in the waveform.The calculation based on the EMD method and Baker’s method identifies the response spectrum of the pulse and the original seismic record,and the comparison is made.The results show that the proposed method has a better agreement with the long-period period of the response spectrum in general,whether it is elastic or inelastic response spectrum.(3)Based on the model by Mavroeidis and using the EMD pulse recognition method in this paper,a pulse analysis algorithm is proposed,which the pulse not only can be adaptively identified,but also can be parsed by the parameters including pulse frequency,pulse shape parameter and impulse oscillation parameter.It is verified that most of the seismic pulses can be resolved by the method in this paper,which provides an effective basis for the correlation analysis between the seismic mechanism and the pulse parameters.