基础沉降的组合预测法

通过对基础沉降的发生过程、特点及灰色Verhulst模型特点的分析,提出可以根据施工过程中的观测资料,运用基于BP神经网络的组合预测模型对不同时刻的基础沉降进行预测;首先分别利用灰色Verhulst模型和BP神经网络模型对基础沉降进行估算,然后利用人工神经网络中的BP神经网络对采用前2种模型所得的结果进行组合预测。计算实例表明,使用该组合预测方法所得到的预测结果比单独使用灰色Verhulst模型或BP神经网络模型所得到的预测结果的总体误差要小,因而该方法是可行的、有效的;可以运用到实际工程中。

Combined Forecasting Method for Foundation Settlement

By analyzing the process and characteristics of foundation settlement and the characteristics of grey Verhulst model, the foundation settlement of various periods after-construction can be predicted according to the observed values during construction period by combined forecasting model, which is based on BP neural network. Firstly, the grey Verhulst model and the artificial neural network(ANN)model are used separately to estimate the foundation settlement; Secondly, BP neural network is employed to forecast the foundation settlement based on the above two estimating results. The results show that the error by this combined forecasting method is smaller than that by grey Verhulst model or ANN model alone. So, this method is feasible and effective, and can be applied to practice.