小波神经网络在群桩基础轴力监测中的应用

Application of Wavelet Neural Network to Axial Force Monitoring of Pile Group Foundation

深水群桩基础是目前大跨径桥梁工程采用的主要基础型式之一。为了研究其荷载传递机理、群桩效应、上部结构—桩—土的共同作用,信息化施工技术已在多座大型桥梁上得到了运用。然而受水文、气象及工程施工等诸多外界因素的干扰,所监测到的轴力时程曲线存在许多突变点,严重干扰了桩基础承载力的分析和预测。为此,提出基于小波神经网络的预测模型,首先采用小波分析对原始监测数据进行去噪,得到反映实际变化的基桩轴力时程曲线,然后分别采用BP神经网络、改进的BP神经网络和径向基函数(RBF)神经网络对其进行预测。研究结果表明:基于小波分析的径向基函数(RBF)神经网络模型预测效果较好。

Deep-water pile group foundation is one of the major foundation types applied in great span bridges.In order to study their load transfer mechanism, pile group effect and interaction among superstructure,piles and soil,information construction technology can be used in lots of large bridges.However,the measured data are usually disturbed by many external factors,such as hydrological,meteorological and construction factors.As a result,there are a lot of breakpoints in the axial force real-time curve which heavily disturb the analysis and prediction of the bearing capacity of pile foundation.Therefore, prediction model based on the wavelet neural network is proposed.Firstly, the wavelet analysis can be used to eliminate the noises from the measured data,so the pile axial force realtime curve that is close to the practical situation can be obtained.Afterwards, the prediction is carried on by BP neural network,improved BP neural network and radial basic function（RBF） neural network.The results show that the prediction effect of RBF neural network is the best.