摘要： 地震活动存在着高度的复杂性和非线性关系，难以挖掘出隐含的相关地震区域，为此提出将极值理论和地震时间序列匹配相结合用于地震趋势研究的基本方法。该方法基于形态的时间-震级二维相似性匹配算法，运用极值理论对相似区域的地震目录进行统计研究，从而验证区域之间存在相关性。对给定区域的历史数据应用该方法，结果证明其能对未来地震趋势做出判断。
关键词:
极值理论,
复发周期,
时间-震级二维相似性,
时间序列匹配,
中强地震
Abstract: It is difficult to dig out the relevant hidden earthquake region because of the high complexity and non-linear relationship in seismic activity. To solve this problem, a new method is used to study the seismic trend combining time sequence similarity matching with Extreme Value Theory(EVT), including morphology-based time sequence-earthquake magnitude Two-Dimensional(2D) similarity matching and statistical research by applying EVT. Some correlation between the regions is verified according to those methods. The effectiveness of the method is verified and the judgment is made by applying it in the historical data.
Key words:
extreme value theory,
recurrence period,
Two-Dimensional(2D) similarity between time-magnitude,
time-sequence matching,
moderate-strong earthquake
中图分类号:
邱剑锋, 谢娟, 李炜, 汪继文. 中强地震的相关性与周期性研究[J]. 计算机工程, 2011, 37(10): 16-18.
QIU Jian-Feng, XIE Juan, LI Wei, HONG Ji-Wen. Research on Correlation and Periodicity of Moderate-strong Earthquake[J]. Computer Engineering, 2011, 37(10): 16-18.