基于BP神经网络的区域短期地震预测模型研究

在详细分析和归纳地震数据及其特征的基础上,提出了可以对地震震级进行量化预测的方法,以江苏省区域为研究对象,采用线性回归和常规BP神经网络方法进行了震级预测,并在总结了这两种方法优缺点的基础上,提出了"线性回归+神经网络"的融合模型。回溯检验结果表明,该模型的预测效果得到显著改善。

Study on Regional Short-term Earthquake Prediction Model Based on BP Neural Network

Based on detailed analysis and induction of the seismic data and their characteristics, this paper puts forward a method for making the quantitative prediction for earthquake magnitude, and taking the region of Jiangsu province as the research object, and makes the magnitude prediction adopting the linear regression analysis and conventional BP neural network, and based on summarizing the advantages and disadvantages of the two methods, puts forward a ＂linear regression ＋ neural network＂ integration model. The results of retrospective inspection show that the predictive effect of this model has been improved significantly.