粉土地震液化等级评判的人工神经网络技术应用

Application Research of Neural Network On The Evaluation of Sandy Loam's Earthquake Liquefaction

据地震液化的实测资料 ,即地震荷载下实测的各项土的物理力学参数 ,利用BP人工神经网络理论 ,建立粉土液化等级评别的反向神经网络模型进行判别 .与现有的方法比较 ,具有通用性、客观性、科学性 .因人工神经网络具有良好的非线性映射能力 ,可以灵活方便地对多成因的复杂的未知数进行高度的建模 .神经网络兼有概念直观 ,公式简洁 。

Based on the paractical date of earthquakes liquefaction and the soil physical parameters that are gained under the earthquake situation,the authors use BP artificial neural network theory to develop a back propagation neural network model of the evaluation of sandy loams earthquake liquefaction.Compared with conventional model,this model is more parctical and can be used,Meanwhile this model has the advantages of dislinct Coneept,brief formula,standavtiyation programcomposition and easy realigation.