摘要
BP算法能够为大坝的变形预报提供较好的预测值,且具有较高的精度。依据BP算法的特性,提出了通过添加学习速率自适应与附加动量项的方法来提高BP算法的运算速度;并对此改进策略进行了试验论证。
BP algorithms can be used for dam deformation forecasting, and can get a well result. In this paper,in order to improve the speed of BP algorithm, give an improve of the algorithm with additional momentum and self-adaptive learning rate, and do a test to prove the right of the improvement.
出处
《城市勘测》
2009年第3期112-115,共4页
Urban Geotechnical Investigation & Surveying
关键词
神经网络
BP神经网络
变形监测
Neural Network
BP Neural Network
deformation forecasting
