东北中强地震发震时间的模式识别预测模型

本文通过对本世纪来东北浅源中强震前地震活动图象的研究,在计算机上建立了“东北浅源中强地震的模式识别综合预测模型”。结果表明:M≥6.0深源地震对东北浅源中强震的发生有重要影响。地震活动频度增加至一定水平,b值下降至一定水平,地震频度和强度随时间的增强,是重要震兆。地震活动的可公度性在识别中也起到了较强作用。本文采用预选特征的方法和标准,加速了建模过程,提高了模型稳定性和效果。控制实验表明结果是稳定的,可用于中强震监测参考。

THE PATTERN RECOGNITION MODEL FOR PREDICTING THE OCCURRENCE TIME OF MODERATELY-STRONG EARTHQUAKES IN NORTHEAST CHINA AREAS

Based on the study of the seismicity patterns prior to shallow moderate events occurred in northeast China since the biginning of the century, the comprehensive prediction model by pattern recognition for shallow moderate earthquakes occurred in northeast China is constructed on the computer. The result shows occurrences of M6.0 events with deep foci can greatly affect the occurrences of shallow moderate events. Significant seismic prcsursors can be obtained when frequency of seismic activities increases and the b-value drops to certain levels, and when the frequency and strength of earthquakes show time-dependent rises. The commensurability of seismicity also plays an important role in the pattern recognition.This paper also illustrates how the process of model construction is quickened, and the stability and effectiveness of the model enhanced through the adoption of the method and standard used to prc-sclect the seismic features. Our controlled experiments show the result is stable and can be used as a reference for the prediction of moderate events.