本文利用模式识别方法对强震孕震过程甲的各种特征进行了选择和分析,检验了诸特征间的相关性并验证了各特征的作用,提出了采用少数特征描述孕震过程和进行分类预报的途径,分析了孕震过程中的典型性质以及典型的时间段,为进一步提高分类效果提供了基础. In this paper, the pattern recognition method is used to select and analyze various features of the seismogenic process of strong earthquakes. The correlation between the features is verified and the function of each feature is verified. A few features are used to describe the seismogenic process and to classify The methods of forecasting, the typical properties during the earthquake seismogenic process and typical time periods are analyzed, which provide a basis for further improving the classification results.