论文部分内容阅读
利用谐小波变换对实际强震记录的时变谱进行估计,并统计分析了与抗震规范相容的3类不同场地上地震波的时变谱特征,分析发现对于较硬场地上的地震波在时域内平稳段较短,下降段衰减较快,而在频域内具有较大的中心频率和较宽的频带;利用均匀调制非平稳随机模型模拟地震波的时变谱,把非线性函数的参数识别问题转化成求解无约束优化问题,利用拟牛顿迭代法求最优解,得到3类不同场地上地震波的均匀调制非平稳随机模型中各谱参数和时间包络中各参数的具体数值,为了定量地确定模拟模型的精度,定义了误差函数,验证了所提时变谱参数识别方法的精度;最后给出了与建筑抗震规范相对应的不同场地、不同烈度下多遇和罕遇地震的谱强度因子.
The time-varying spectrum of actual strong earthquake records was estimated by using the harmonic wavelet transform, and the time-varying spectrum characteristics of the seismic waves on three different sites that were compatible with the seismic code were statistically analyzed. It was found that for the seismic wave in the hard field in the time domain The stationary section is short, the descending section decays rapidly, but has a larger center frequency and wider frequency band in the frequency domain. By using the uniformly modulated nonstationary stochastic model to simulate the time-varying spectrum of seismic waves, the parameter identification problem of the nonlinear function is transformed In order to solve the unconstrained optimization problem, the optimal solution is obtained by using the quasi-Newton iteration method, and the specific values of each parameter in each spectrum parameter and time envelope of the uniformly modulated non-stationary stochastic model of the seismic on three different sites are obtained. In order to quantitatively determine The accuracy of the model was simulated and the error function was defined to verify the accuracy of the proposed time-varying spectral parameter identification method. Finally, the spectral intensity factors of different occasions and rare earthquakes under different intensities were given. .