考虑超随机特性的人工地震动合成

The synthesis of man-made ground motion considering the super-randomness

文中根据自相关函数随机噪声部分的2个重要性质提出了自相关函数随机噪声部分的迭代生成算法,并给出了以确定性自相关函数为目标函数,考虑随机噪声部分影响的人工合成地震动方法。本文的研究结果表明地震动加速度的平方和(SS)比峰值加速度(PGA)更好地量度了地震地面运动强度。文中用蒙特卡罗方法分析了自相关函数噪声项对人工合成地震动幅值和反应谱的影响:由于自相关函数的随机噪声项对地震动总功率或平方和的贡献为0,因此随机噪声项对合成地震动幅值的影响不大,但由于自相关函数随机噪声项明显改变了地震动功率谱在频域上的分布特征,使原本光滑的平均反应谱出现了明显的波动,从而在一定程度上提高了平均反应谱的峰值,使自相关函数噪声部分对平均反应谱也有一些影响。

According to the two important properties of the autocorrelation function, the paper proposes a method of creating the random noise in autocorrelation function through iterations and then gives a method to synthesize manmade seismic wave whose objective function is the autocorrelation function or the corresponding power spectra in consideration of so called super-randomness of ground motion caused by random noise in autocorrelation. It is indicated that the SS （ sum of squares） of ground motion is more adaptable measurement of the ground motion severity rather than the peak ground acceleration （PGA） of ground motion while whose main phase can be regarded as a stationary random process. The Monte Carlo method was used to analyze the effect of the random noise part in autocorrelation on the amplitude of man-made seismic wave and on the response spectra as well. Due to the zero contribution of the noise part of auto-correlation to the total power of seismic wave, the effect of the noise part on the man-made seismic wave＇s amplitude is insignificant. But because the noise part greatly changes the frequent-domain distribution of the power spectrum, the random part in autocorrelation function leads to some effect to certain extend on the response spectra.