摘要
冻土区公路路基沉降影响因素众多,预测难度大,针对目前传统的预测方法所面临的困难,应用人工神经网络理论,建立预估模型。分别对影响沉降的因素诸如气温,降水,交通量荷载进行预测,将前3个网络的影响因素变量的预测结果(即输出值)作为预测沉降的网络输入训练网络,得到冻土区公路路基沉降的预测值。使用这种方法,可以代替常规的力学数学方法,具有一定的适用性。
There are lots of factors which will affect the subgrade deformation in permafrost area.And it is difficult to anticipate because of complex background . To aim directly at the existing deficiency of traditional prediction method for subgrade settlement, the prediction model based on the theory of artificial neural network (ANN) is developed. There are three nets including temperature, precipitation and AADT, and the output results of the above three nets will be as the input vectors for the net to anticipate the settlement.After training, the prediction of subgrade settlement can be gotten.This approach can replace the commonly used mechanics math method to calculate the settlement and is feasible in practice.
出处
《公路交通科技》
CAS
CSCD
北大核心
2005年第11期42-44,共3页
Journal of Highway and Transportation Research and Development
基金
交通部西部交通建设科技项目(2002-318-795-08)
关键词
神经网络
BP算法
路基沉降
冻土区公路
Neural network
BP algorithm
Subgrade settlement
Permafrost area road