砂土地震液化预测的Bayes判别模型及其应用

Bayes discriminant analysis model for seismic sand liquefaction

基于Bayes判别分析理论,在分析影响砂土液化因素的基础上,选取影响地震砂土液化因素的8个实测因子(地震烈度I、震中距L、地下水位dw、砂层埋藏深度ds、标贯击数N63.5、砂土颗粒平均粒径d50、不均匀系数Cu和剪应力比τd/σv)作为砂土地震液化判识指标,以唐山大地震中21组典型的砂土地震液化资料为学习样本,建立了唐山地区砂土地震液化预测的Bayes判别分析模型,并选用其他9组待检样本对判别模型进行验证。结果表明,Bayes判别分析模型预测性能良好,具有预测准确度高、回判估计的误判率低等特点,是解决砂土液化预测问题的有效方法之一。

Based on the principle of Bayes discriminant analysis and the factors influencing the sand liquefaction,the model was applied to the prediction of sand liquefaction with eight factors listed as following: seismic intensity,epicenteral distance,groundwater level,sand depth,blow number of standard penetration test,mean granular diameter,coefficient of nonuniformity,ratio of shearing stress to effective overburden stress. With 21 groups of typical sandy soil seismic liquefaction data during Tangshan earthquake as samples,Bayes discriminant analysis model of predicting earthquake liquefaction of sandy soil in Tangshan area is established. Then the model is applied to predict the sand liquefaction through other 9 samples. The results show that the prediction Model of Bayes discriminant analysis has excellent performances with high prediction accuracy. It is an efficient method for the prediction of sand liquefaction and could be applied to practical engineering.