基于RBF神经网络的砂土液化预测

Application of RBF Neural Network to Prediction of Sands Liquefaction Potential

通过分析砂土液化成因及其影响因素,建立了砂土液化预测RBF网络模型,并与BP网络预测模型进行比较.测试结果表明,应用RBF网络模型对砂土液化进行预测,预测效果好,识别精度高.

By analyzing the causes and contributing influences of sands liquefaction potential, the RBF neural network model for prediction is set up and compared with the BP network. The testing results show that RBF network model for predicting sands liquefaction potential can become a new effective assessment model with better prediction results and higher recognition precision.