基于小波—神经网络的深基坑沉降预测研究

Research on settlement prediction of deep pit based on WRPM

将小波分析与径向基函数(简称RBF)神经网络相结合,构建了基于智能算法的深基坑沉降预测模型(简称WRPM)。根据深基坑工程的施工特点,借助WRPM分析了影响沉降的主要因素,研究了基坑沉降机理,提取了沉降的真实信号并对沉降变形进行了预测。工程实例分析表明,WRPM模型用于深基坑沉降预测,具有精度高、泛化能力强等特点,能够为深基坑工程的安全施工提供依据。

This paper combined wavelet analysis and radial basis function neural network（RBF）,and established a deep pit settlement prediction model based on the intelligent algorithm（namely WRPM）.According to the characteristics of deep pit engineering,the main factors affecting the deep pit settlement were analyzed by the WRPM,and the mechanism of deep pit settlement was researched.The true signals of settlement were extracted to predict the settlement.The example show that WRPM has the high precision and generalization ability for the deep pit settlement prediction.The model can provide the basis for the safety construction of deep pit engineering.