论文部分内容阅读
以现场实测资料为依据,建立了抗滑结构位移预测的神经网络模型,分别用于抗滑桩顶位移预测和抗滑桩沉降预测,并与灰色模型的预测结果进行了比较,结果表明,神经网络的预测结果更接近于实际值,从而验证了神经网络用于抗滑结构位移预测的可行性。神经网络收敛快、拟合效果好、泛化能力强、预测精度高,是抗滑结构位移预测的有效方法。
Based on the field data, a neural network model of displacement prediction of anti-skid structure is established, which is respectively used for prediction of displacement of anti-slide pile and prediction of anti-slide pile settlement, and compared with gray model prediction results. The results show that nerve The predicted result of the network is closer to the actual value, which verifies the feasibility of using neural network to predict the displacement of anti-slip structure. Neural network convergence fast, good fitting effect, generalization ability, high prediction accuracy, is an effective method to predict the displacement of anti-sliding structure.