多参量动态综合预测模型  

Time Dependent Synthetic Prediction Model with Parameters

作　　者：戴英华[1] 刘永强[1] 孙佩卿[1] 

机构地区：[1]河北省地震局,河北机电学院

出　　处：《中国地震》1996年第1期26-34,共9页Earthquake Research in China

基　　金：地震联合基金会

摘　　要：现有的综合预测模型，大多未考虑系统的时变性和各种异常在震前所显示的不同持续时间。本文针对地震过程的不可逆性与孕育过程中的时变性，由可容纳不同异常持续时间的多参量动态预测综合模型，依据适合于不可逆过程的ＡＲＭＡ和ＶＡＲ参量预测模型，以华北北部区域为例，探讨了进行未来强震危险性预测的途径。最后，对影响预测精度的因素进行了分析和讨论。For present synthetic models, time dependent characteristics of the system and different anomalous periods of various precursors shown before strong earthquakes have been not considered. In this paper, to counter the non-reversibility and time dependent characteristics of the preparatory process of a strong earthqunke, a time dependent prediction model with parameters to hold different anomalous periods have been taken basing on ARMA and VAR parameter prediction model which is suitable to the nonreversible process. As an example, seismic hazard in North China has been shown to explore the way to estimate seismic hazards in future. Finally, the factors to affect on prediction precision have been analyzed and discussed.

关 键 词：地震预报 多参量动态 综合预测模型 

分 类 号：P315.7[天文地球—地震学]