深基坑变形的神经网络预测分析

Prediction of Deformation in Deep Cut Based on Neural Network

采用神经网络理论建立了深基坑开挖引起的周围地表移动变形问题分析的非线性时间序列神经网络模型,并用该模型对深基坑开挖引起的周围地表移动变形工程实例进行了具体的研究分析。将预测结果与现场实测资料比较,二者吻合得较好。结果表明,建立的非线性时间序列神经网络模型能够有效的表达深基坑变形规律并能对未来变形量进行适时预测。

A nonlinear time series neural network model that predicts ground surface displacement caused by deep cut is established based on the theory of neural network. The model is developed and applied in several cases to predicting displacement and deformation of ground surfaces that are adjacent to deep cut. And the agreement of modeling results with the on-site monitoring data indicates that this model is satisfactory and effective in expressing the deformation rules and in predicting a further deformation of the deep cut.