【摘要】 2012年9月7日我国西南云贵交界彝良-威宁地区连续发生两次MS5.5和MS5.6中强地震,其时间间隔短,震中位置接近,震级相差很小,为中强地震地震动的随机特性分析研究提供了难得的强震动数据.本文从集系与样本函数两方面考虑,通过加速度时程的相关函数研究了地震动随机过程的相关特性.结果表明:两次地震集系的相关矩阵均为随机相关,互相关矩阵较自相关矩阵表现出更强的随机性;两次地震在同一个台站记录的加速度时程所构成的集系,其相关性较两次地震单独构成的集系好;样本函数的相关函数均呈现不规则震荡衰减的特点,且两个采样值间时滞越大,其样本函数相关性越小.相比大震(如我国台湾地区的集集地震),中强地震地震动记录的相关矩阵及分解后的自相关函数均表现出不同的特点.由于这两次地震的地震动幅值相比大震较小,土层非线性影响较弱,故本文结果为认识地震动的随机特性提供了新的参考依据.更多还原

【Abstract】 On 7September 2012,two earthquakes took place consecutively at the boundary between Yiliang county in Yunnan Province and Weining county in Guizhou Province,southwestern China.They are characterized by a short timeinterval,close epicentral position and small difference in the earthquake magnitude,which provides a number of valuable strong ground motion recordings for analyses and researches on random characteristics of ground motions.In the present paper,the correlative characteristics of seismic stochastic process through correlation functions of acceleration histories are studied from the aspect of the set systems and sample functions.The results show that,correlation matrices of set systems for ground motions during the two earthquakes are random.Comparison with autocorrelation matrices indicates that cross-correlation matrices show greater randomness.Set systems that are composed of acceleration histories for the two earthquakes with the same station are correlated better than those for individual earthquake.Correlation functions of sample functions both show irregular fluctuant attenuation characteristics;correlation of sample functions becomes poorer with time difference of the sampling values increasing.In comparison with great earthquake(such as the Chi-Chi earthquake),correlation matrices and autocorrelation functions after decomposition of the moderate ground motion recordings both exhibit different characteristics.Because both two earthquakes have smaller ground motion amplitudes and weaker soil nonlinear effect,so the results provide new references for understanding the random characteristics of ground motions.更多还原