BP和Elman神经网络在砂土液化预测中的研究  

Study on forecasting of sand liquefaction by using BP neural and Elamn neural networks

作　　者：高宗军[1] 付青[1] 郑秋霞[1] 王世臣[1] 许传杰[1] 董红志[1] 

机构地区：[1]山东省沉积成矿作用与沉积矿产重点实验室/山东科技大学地质科学与工程学院,山东青岛266590

出　　处：《中国安全生产科学技术》2013年第6期58-62,共5页Journal of Safety Science and Technology

摘　　要：基于砂土液化的影响因素具有非线性关系,而神经网络模型能够逼近任意非线性函数和适合于动态系统辨识的特性,分别建立输入层为4,隐含层神经元为2,输出层为1的三层BP神经网络和Elman网络,并且通过matlab软件运算,实例比较得出Elman模型比BP模型收敛速度快、精度高,在砂土液化的预测中效果更好。Based on the condition that the influencing factors of sand liquefaction have a non-linear relationship,while the neural network model can simulate any nonlinear function and is suitable for dynamic system recognition,in this paper,a three-layer BP neural network and Elman network with 4 input layer neuron,2 hidden layer neuron and 1output layer neuron were established respectively.What＇s more,by the way of practical examples simulations,Elamn neural networks haa a faster convergence speed,a higher accuracy and a better effect than Elamn neural networks.

关 键 词：砂土液化 BP神经网络 ELMAN神经网络 MATLAB软件 

分 类 号：P42[天文地球—大气科学及气象学]