深基坑支护结构位移的神经网络预测

Artificial Neural Network Prediction for Displacement of Support Structure in Deep Foundation Pit

针对深基坑系统的复杂性和变形非线性,将人工神经网络技术引入其中。在分析讨论了人工神经网络中应用最为广泛的BP网络基本原理的基础上,建立了基坑变形预测的神经网络模型,并应用Bayesian方法对实例加以论证。研究表明,神经网络是解决基坑变形预测的有效方法之一。

In allusion to the complexity of the deep foundation system and the nonlinear of deformation, the artificial neural network technology is cited. Based on the analysis of the fundamenta! principle of BP （Back Propagation） neural network which is the most popular one, the neural network model for predicting the foundation pit deformation is established. And the Bayesian method is applied to demonstrate the examples. The research indicates that the artificial neural network is one of the valid and useful methods for the prediction of the foundation pit deformation.