砂土液化预测的人工神经网络模型

Artificial Neural Network Model for Prediction of Liquefaction of Sandy Soil

摘要: 本文根据人工神经网络的一典型模型—反向传播模型,以及地震荷载下的各项土的物理—力学参数,建立了土液化类型的神经网络数学模型。研究表明,人工神经网络方法性能良好,可望成为砂土液化预测的有效手段。

Abstract: A typical artificial neural network model-back-propagation model was presented for prediction on the soil liquefaction type based on the physical parameters of soils under earthquake. Results showed that the neural network approach is good, and might be referred as an effective technique for prediction of the liquefaction of sandy soil.