The Influence Of SNR Selection On Response Spectrum Calibration In Loess Site

The strong earthquake record is the basis of studying the response spectrum.The influence of the selection of SNR on the design response spectrum is an important issue in the field of seismic engineering during the processing of strong earthquake records.The influence of the selection of cutoff frequencies with different SNR on the acceleration response spectrum and its calibration is studied in this paper.The data were based on the strong earthquake records collected from loess sites in the last ten years.The influence of high pass cutoff frequency and low pass cutoff frequency on the ground motion parameters and the response spectrum of displacement,velocity and acceleration is systematically analyzed.At the same time,the acceleration response spectra after different filtering processing are calibrated.After calibration,the design acceleration response spectrum and characteristic parameters were compared and analyzed.A new calibration method of the design reaction spectrum is also proposed.The main results of the thesis are as follows:1?The noise sources in analog and digital strong earthquake records are analyzed.The theory of causality and non-causality in filters is studied in detail.The results show that causal filtering always causes the phase difference of ground motion before and after filtering,and non-causal filtering is better than causal filtering in eliminating phase distortion and broadband.The realization of non-causal filtering requires adding zero to the original acceleration records.The purpose of adding zero is to avoid drift and tail warping in the displacement history obtained by acceleration integration.Use cosine taper to add a smooth zero to the filling section and to record the transition bands at both ends,and the plus zero part should be retained in the calculation of velocity and displacement.The research shows that the Butterworth filter has the best stability in each filtering order,so this paper adopts the four-order Butterworth bandpass filter commonly used at home and abroad.2?The processing method of strong earthquake data is summarized systematically,and the filtering processing flow is given.The high-pass cut-off frequency of filtering is selected according to multiple criteria such as single inflection point source model,double Angle source model,source spectrum theory and recorded SNR.The strong seismic records of loess and non-loess sites were used to verify the results.The results show that multiple criteria such as source spectrum model,source spectrum theory and recorded signal-to-noise ratio are used to determine the high-pass cutoff frequency of the filtering,and the subjective influence is reduced to the minimum,which provides a reference process for analyzing the influence of filtering based on different signal-tonoise ratio on the response spectrum and the selection standard of cutoff frequency.3?The influence of high-pass-off frequency on the design reaction spectrum is studied.Based on the filtering processing of strong earthquake records with different high-pass cutoff frequencies determined by different signal-to-noise ratios,the influence of different high-pass cutoff frequencies on relevant ground motion parameters and response spectrum is analyzed.The results showed that PGA,PGV and PGD tended to increase with the increase of high pass cutoff frequency.Compared with PGA and PGV,PGD increased more and was more sensitive to the selection of high pass cutoff frequency.The effect of the high pass cutoff frequency on the response spectrum is also the same.The effect of the high pass cutoff frequency on the acceleration and velocity response spectrum is not obvious,but the effect on the displacement response spectrum is very significant.Furthermore,we calibrate the acceleration response spectra after different filtering processing,and find that the calibration parameters have little difference,and the shape of the designed response spectra is basically the same.This indicates that the influence of high-pass cutoff frequency on the acceleration response spectrum and its calibration is small,but the influence on the displacement and the displacement response spectrum is very significant,so the influence of high-pass cutoff frequency should be taken into account in the study of the displacement response spectrum.4?The effect of low pass cutoff frequency on the design reaction spectrum was studied.By filtering the original records with different low-pass cutoff frequencies,it is found that the absolute value of ground motion parameters does not change much when the low-pass cutoff frequency increases from 20 Hz to 90 Hz,especially the values of PGD and PGV almost do not change,and the variation range of PGA is within 5%.The response spectra of acceleration,velocity and displacement were calculated from the processed data.From the obtained response spectra,the influence of low-pass cutoff frequency on the response spectra was not obvious.The calibration results show that the low-pass cutoff frequency has little effect on the calibration of the acceleration response spectrum,and the calibration parameters are basically the same.This indicates that the parameters and response spectrum of ground motion are not sensitive to the selection of low-pass cutoff frequency compared with the high-pass cutoff frequency.5?A new calibration method of design reaction spectrum is proposed.PSO is introduced into the calibration of the design reaction spectrum.Particle swarm optimization algorithm is the result of the study of birds feed on behavior of swarm intelligence is to simulate set up by a global optimization algorithm,to deal with the problem of multiple parameters of nonlinear function optimization has good performance,in order to have better convergence accuracy and faster convergence speed,this paper constructed with compression factor of particle swarm optimization(pso)algorithm,used to calibrate the design response spectrum.By using this method,the first inflection point period,characteristic period,platform value and attenuation index can be obtained to characterize the design reaction spectrum.In this paper,the calibration of El Centro acceleration time-history response spectrum is taken as an example,Newmark three-parameter method,two-parameter method and differential evolution algorithm are used to calibrate it.The characteristic parameters and calculation accuracy of the four calibration methods are compared and analyzed.The example shows that the improved particle swarm optimization algorithm has higher precision and the designed response spectrum can reflect the characteristics of the seismic response spectrum.