汉十高铁线下工程沉降预测新陈代谢GM-BP组合模型研究
|
摘要
提出了一种新陈代谢GM(1,1)模型与BP神经网络模型的组合预测模型,给出了组合模型的基本思路,通过BP神经网络计算而来的误差预测值对新陈代谢GM(1,1)模型的预测值进行改正,并将新陈代谢GM-BP组合预测模型应用到新建铁路武汉至十堰铁路、孝感至十堰段的高铁陆地桥墩沉降预测中,结果表明:新陈代谢GM-BP组合预测模型较现有的新陈代谢GM(1,1)模型具有较高的精度。
This paper proposes a combined forecasting model of metabolic GM(1,1) model and BP neural network model,and gives the basic idea of the combined model. The error prediction value calculated by BP neural network is used for metabolic GM(1,1). The predicted values of the model is corrected by the error prediction,and the metabolic GM-BP combined forecasting model was applied to the prediction of the settlement of high-speed railway land piers from the Xiaogan-Shiyan section of the newly built railways from Wuhan to Shiyan. The results show that the metabolic GM-BP combination forecasting model is superior to the existing ones. The metabolic GM(1,1) model has high accuracy.
This paper proposes a combined forecasting model of metabolic GM(1,1) model and BP neural network model,and gives the basic idea of the combined model. The error prediction value calculated by BP neural network is used for metabolic GM(1,1). The predicted values of the model is corrected by the error prediction,and the metabolic GM-BP combined forecasting model was applied to the prediction of the settlement of high-speed railway land piers from the Xiaogan-Shiyan section of the newly built railways from Wuhan to Shiyan. The results show that the metabolic GM-BP combination forecasting model is superior to the existing ones. The metabolic GM(1,1) model has high accuracy.
引文
[1] 李涛,花向红,刘金标等. 基于新陈代谢的动态灰色模型及其应用[J]. 城市勘测,2008(4):135～137.
[2] 谢波,汪荃. 最小一乘法GM(1,1)在变形监测中的应用[J]. 城市勘测,2013(6):150～152.
[3] 林淼. 一种改进的高铁桥沉降变形预报方法研究[J]. 城市勘测,2017(6):135～138.
[4] 吴乃龙,周永领. 基于BP神经网络改进算法在地铁隧道施工中的沉降预测[J]. 城市勘测,2015(3):148～150+157.
[5] 黄声享,尹晖,蒋征. 变形监测数据处理[M]. 武汉:武汉大学出版社,2010.
[6] 覃劭峰. 基于GM(1,1)-ARIMA最优组合的大坝预测模型[J]. 测绘通报,2014(S2):66～69.
[7] 唐小我. 最优组合预测方法及其应用[J]. 数理统计与管理,1992(1):31～35.
[8] 翟信德,高飞,徐文兵. 新陈代谢GM(1,1)模型在建筑物沉降预测中的应用研究[J]. 城市勘测,2008(3):136～138.
[9] 曾贤敏,黄腾,李桂华等. 新陈代谢GM(1,1)模型在大坝边坡沉降监测中的应用[J]. 测绘工程,2015(8):53～56.
[2] 谢波,汪荃. 最小一乘法GM(1,1)在变形监测中的应用[J]. 城市勘测,2013(6):150～152.
[3] 林淼. 一种改进的高铁桥沉降变形预报方法研究[J]. 城市勘测,2017(6):135～138.
[4] 吴乃龙,周永领. 基于BP神经网络改进算法在地铁隧道施工中的沉降预测[J]. 城市勘测,2015(3):148～150+157.
[5] 黄声享,尹晖,蒋征. 变形监测数据处理[M]. 武汉:武汉大学出版社,2010.
[6] 覃劭峰. 基于GM(1,1)-ARIMA最优组合的大坝预测模型[J]. 测绘通报,2014(S2):66～69.
[7] 唐小我. 最优组合预测方法及其应用[J]. 数理统计与管理,1992(1):31～35.
[8] 翟信德,高飞,徐文兵. 新陈代谢GM(1,1)模型在建筑物沉降预测中的应用研究[J]. 城市勘测,2008(3):136～138.
[9] 曾贤敏,黄腾,李桂华等. 新陈代谢GM(1,1)模型在大坝边坡沉降监测中的应用[J]. 测绘工程,2015(8):53～56.