摘要
应用“灰色控制系统”理论,选取1980年1月至1988年12月日本的地震序列资料,将(6.0,7.0]级地震作为样本,建立了预报地震发震时刻的动态模型。选取了最佳模型对日本未来发生中强地震的时间进行了预测。从函数变换的观点,对GM(1,1)模型进行了广义解释,指出序列建模必须从满足光滑度的时刻计起。大量计算表明,用“足够小量”样本建模比大量样本建模拟合与外推精度要高,并从信息论角度进行了剖析。
A dynamic model for forecasting the occurrence time of earthquake is established by using “Grey controlSystem”theory,and the earthquake sequence data in Japan from ’980~1988 are choosen to take earthquakeswith magnitude(6.0,7.0] as samples. A optimal model is adopted to forecast the occurrence time of mode-rate strong earthquake in Japan. The model GM (1,1)is explained generally on the view of function transfor-mation,and it is shown that the degree of slick must be satisfied in order to establish the sequence model.Nu-merical results show that the accuracy of fitting and extrapolation of the sequence modelling on″ adequate mi-nority″samples is higher than on mass samples,and the paper analyzes the accuracy based on information the-ory.
出处
《山西地震》
1994年第2期8-10,共3页
Earthquake Research in Shanxi
关键词
中强地震
灰色预测
地震
Japan miderate strong earthquake Grey forecast