摘要
为解决神经网络在沉降预报应用中的局限性,结合灰色理论等维信息策略和BP神经网络建模思想,利用数据序列本身构建训练样本,建立等维BP神经网络预报模型,并利用数学工具MATLAB编程实现,进行实际的变形预测,通过与灰色理论GM(1,1)模型的预报效果进行比较,表明该等维BP网络模型具有更高的预报精度,可以达到更好的预报效果。
Based on equal dimention strategies of grey theory and BP neural network, this paper constructs the training sample with the data sequence itself, establishes the equal dimentional BP neural network, and applies this model in the practical deformation forecast with the mathematical instrument—— MATLAB. Compared with the grey theory GM（1,1） model, this new equal dimentional BP neural network has higher forecast precision, and may achieve better forecast effect.
出处
《测绘工程》
CSCD
2008年第3期10-12,15,共4页
Engineering of Surveying and Mapping
关键词
等维
BP神经网络
变形分析
沉降预报
equal dimension
BP neural network
deformation monitoring
sedimentation forecast
