It is the key of seismic risk analysis to reasonably predict the possible ground motion of engineering site in the future,stochastic modeling of ground motion is a simple tool to predict ground shaking level for future earthquake and less time consuming than physics-based deterministic modeling.Based on the modulated filtered white noise model and NGA-West2 strong motion database,the research of the generating ground motion acceleration time histories under specific conditions has been carried out.The synthetic artificial ground motion time history captures the main characteristics of the real ground motion,which can provide help for the nonlinear dynamic analysis of the structure.In addition,the sensitivity analysis of the statistical data in the random synthesis of ground motion is also carried out.This article mainly completed the following work:(1)Based on 2571 ground motion records in NGA-West2 database,the non-stationary parameters in time domain and frequency domain of each ground motion time history are identified,and the empirical prediction equations of six characteristic parameters in standard normal space are constructed.Based on the regression results,the correlation between the six characteristic parameters and the magnitude,fault distance and site conditions,as well as the correlation between the six characteristic parameters are discussed.(2)With the help of the stochastic synthetic method of broadband ground motion,the broadband ground motion with setting magnitude,fault distance and site condition is generated.The synthetic artificial ground motion can accurately reflect the temporal nonstationarity and the spectral nonstationarity of the real ground motion,which can provide help for the seismic design and analysis of the structure.(3)Based on the synthesized ground motion parameters,the GMPEs of PGA and PSA with different periods are obtained,which is compared with other NGA-West2 GMPEs.It is found that:for large earthquakes with magnitude more than 6,the GMPEs simulated in this paper fits well with other GMPEs that have been applied in practice.(4)In this paper,the influence of statistical data of different magnitudes and fault distances selection range on the amplitude and error of synthetic earthquake is studied by using this stochastic synthetic method.The results show that the amplitude of synthetic ground motion is less sensitive to the selected range of statistical data M,and more sensitive to the selected range of statistical data Rrup;The larger the magnitude of the target ground motion or the smaller the fault distance,the lower the sensitivity of the amplitude of the composite ground motion to the statistical data;The wider the range of statistical data selection,the greater the error in simulation. |