砂土液化预测的Fisher判别分析模型及应用

Fisher discriminant analysis model and its application to sand liquefaction prediction

针对砂土地震液化预测问题,基于Fisher判别分析原理(FDA),选用平均粒径、不均匀系数、标贯击数、地下水水位、砂层埋深、剪应力与有效上覆应力比、地震烈度、震中距等8个实测指标为判别因子,建立Fisher线性判别函数模型,对砂土液化进行预测。研究结果表明,Fisher判别分析结果与神经网络输出结果一致,优于规范法和Seed法判别结果,验证了该模型的合理性和可靠性。运用该模型进行判别分析,简易方便,分类效率高,对砂土液化判别快速、有效,模型适用性强,具有一定的工程应用前景。

For the prediction of seismic liquefaction of sand,Fisher discriminant model is established to the prediction of sand liquefaction based on the principle of Fisher discriminant analysis theory with eight factors listed as follows： mean diameter,coefficient of non-uniformity,blow number of standard penetration test,underground water depth,sand depth,ratio of shearing stress to effective overburden stress,seismic intensity,epicenteral distance.Through computing practical examples and assessing the model,the Fisher discriminant analysis model is manifested to be of rationality and reliability and the model identification results are consistent with the neural network output,and are better than the discriminant results with the specification method and Seed method.It is easy and efficient to make discriminant analysis using this model and it can provide efficient means to predict sand liquefaction.