基于BP神经网络的软土路基施工期沉降预测与仿真
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
在高速公路软基沉降预测与施工控制中,把现场填土厚度、时间及实测沉降量信息作为学习样本,经过BP神经网络训练后,建立了动态仿真预测模型,用该模型预测当前及下一级填土后的沉降量,判断下一级填土是否可行及对路堤稳定性的影响,以实现对软基沉降及工程施工的事前仿真控制,避免工程事故的发生。实际工程应用表明,该模型的预测精度满足工程建设的需要。
In the prediction and control of settlement of highway soft foundation,taking the field measuring information of the thickness of roadbed construction,time,and settlements as studying samples,simulated prediction model which varies with time based on BP neural network is established.The model can predict the present and next roadbed construction settlement and thus can also predict whether the next roadbed construction is feasible or not and whether the stability of the roadbed is influential or not.The simulated control of settlement of soft foundation before roadbed construction has been done to avoid engineering accidents.Application in highway construction shows that the prediction accuracy of the model can meet the needs of engineering construction and the method is feasible.
引文
[1]李孝安.神经网络与神经计算机导论[M].西安:西北工业大学出版社,1994.
    [2]LIUYong-jian.Prediction of Final Settlement of Soft Groud for Expressway by Using Artificial[J].Journal of Highway and TransportationResearch and Development,2000,17(16):15-18.
    [3]飞思科技产品研发中心.MATLAB6.5辅助神经网络分析与设计[M].北京:电子工业出版社,2003.
    [4]王伟.人工神经网络原理——入门与提高[M].北京:北京航空航天大学出版社,1995.
    [5]温岩,乔兵,李忠伟,等.基于MATLAB工具箱的神经网络在地震预报中的应用[J].东北地震研究,2001,17(3):36-14.
    [6]杨涛,李国维,樊琨.基于人工神经网络的公路软基沉降预测模型[J].上海理工大学学报,2003,25(2):117-120.

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心