摘要
深基坑变形是评价深基坑开挖过程中安全性的一个重要参数,对其进行精确预测是一个亟待解决的技术难题.为了更加准确地预测深基坑的变形,采用有动量的梯度下降算法,将现场的监测数据作为神经网络的输入参数,建立BP神经网络模型进行深基坑的变形预测.结果表明,模型的预测值与实测值之间的误差在5%以内,满足实际工程的要求.所建模型能够兼顾精度和效率,便于程序实现,能为深基坑的变形预测分析提供有效工具.
The deformation of deep foundation pits is a key parameter to evaluate their safety during the excavation, and precise forecast is an urgent technical problem. To obtain a more accurate predicted deformation, gradient descent algorithm with momentum is employed in the BP neural network model and the field monitoring data is used as input parameters for the neural network. The results show that the error between the predicted value and the measured value is within 5 %, which can satisfy the practical engineering requirements. Accurate and effective, the model is easy to program, thus providing an efficient tool for the deformation prediction of deep foundation pits.
出处
《南京工程学院学报(自然科学版)》
2016年第3期17-22,共6页
Journal of Nanjing Institute of Technology(Natural Science Edition)
基金
高等学校博士学科点专项科研基金项目(20133204120015)
江苏省高等院校自然科学基金项目(12KJB560003)
关键词
深基坑
变形预测
BP神经网络
MATLAB
deep foundation pit
deformation
back propagation neural network
Matlab