开挖降水条件下基坑周围建筑物沉降预测研究

Research on settlement prediction of foundation pit surrounding building in excavation and dewatering conditions

利用Gompertz曲线模型、灰色理论模型和BP神经网络模型等单一预测模型对基坑周围建筑物沉降进行预测,之后将结果与三者优化组合模型的预测结果进行比较分析,结果表明:单一模型预测结果的精度比三者优化组合模型的精度较低,而其中通过最优加权法组合的模型预测精度最高。

This paper made prediction to foundation pit surrounding building settlement using the Gompertz curve model,grey theory model and BP neural network model and other single forecasting model,later made comparative analysis on results and prediction results of three portfolio optimization model,the results showed that the results accuracy of single predict model was lower than three portfolio optimization model,and the forecast accuracy of optimal weighted combination model was the highest.