临近基坑建筑物沉降神经网络动态预报

摘要

A dynamical predication model for building subsidence course in deep excavation and an effective predicting method is put forward by improved artificial neural networks combining BP neural network with genetic algorithm. The result indicates that the model proposed here has fast approximation and high precision, the predicted values agree well with the measured ones. The proposed method is a useful tool for deformation prediction.%将BP神经网络与遗传算法结合,建立了建筑物沉降的动态预报模型.实例分析表明所建立的模型预测精度较高,预报值与实测值吻合较好,该方法对建筑物沉降的实时预报有一定的实用性.