欢迎访问《地震》，
地震

地震 ›› 2000, Vol. 20 ›› Issue (1): 10-16.

• • 上一篇    下一篇

人工神经网络在地震中期预报中的应用

王炜1, 宋先月1, 王琤琤2   

  1. l.上海市地震局、上海 200062;
    2.江苏省地震局．江苏南京 210014
  • 收稿日期:1999-05-26 修回日期:1999-06-26 出版日期:2000-01-31 发布日期:2022-09-26
  • 作者简介:王炜(1947一),男,研究员,主要从事地震学、地震顶报、专家系统、人工神经网络等研究
  • 基金资助:
    中国地震局“九五”科研攻关项目(95-04)

The application of artifieial neural networksto mid-term earthquake prediction

WANG Wei1, SONG Xian-yue1, WANG Zheng zheng2   

  1. 1. Selsmologieal Bureau of Shanghai Munieipality,Shanghai 200062,China;
    2. Seismologeal Bureau of Jiangsu Provinee, Nanjing 210014,China
  • Received:1999-05-26 Revised:1999-06-26 Online:2000-01-31 Published:2022-09-26

PDF (PC)

113

摘要： 将BP神经网络用于地震中期预报。使用-些常用的地震学指标作为神经网络的输人,而将BP神经网络的输出作为表征地震活动增强的特征参数W,,并将其用于华北地区进行空间扫描。结果表明,中强地震前1-3年未来震中周围通常出现明显的Wl值中期异常区,该方法具有较好的中期预报效果。

关键词: BP神经网络, 非线性关系, 地震活动增强, 地震中期预报

Abstract: BP neural network is used in mid-term earthquake predietion.Some usual predietion parameters of seismology are used as input units of neural networks.And the output urits of neural networks are called as the characteristic parameter W1 deserlbing enhancement of seismieity.It is used for spatial seanning in North China.The result shows that the mid-term anomaly zone of W1-value usually appears obviously around the epleenter about l～3 years before future eearthquakes with good predietion effeet.

Key words: BP neural network, Nonlinear relationship, Enhaneement of selsmieity, Mid-term earthquake Predzetion

中图分类号: