摘要
为充分利用铁路构筑物沉降数据表现出来的空间趋势性与相关性,提高预测结果的整体稳健性和准确性,提出一种新的思路:先对构筑物沉降数据序列进行分析,若其表现出较强的空间相关性,则可采用时空序列模型对构筑物沉降数据序列进行建模预测。以江苏省境内某铁路一处3孔箱形桥为例进行验证,该箱形桥上布设有6个沉降监测点,通过分析,监测点沉降数据序列的相关性系数高达0.99,说明该箱形桥的沉降表现出很强的空间趋势性和相关性,故采用时空序列模型对该箱形桥的6个沉降监测点进行建模和沉降预测,并将预测结果与实际监测值进行对比。对比结果表明,时空序列模型预测结果的4种误差指标均小于一般时间序列模型,证明该方法的预测结果更具稳健性和准确性。
To make full use of the spatial trends and correlations shown by the settlement data of railway structures and to improve the reliability and accuracy of the settlement forecasting,this paper presents a new method of modeling railway settlement data.Firstly,the original railway settlement data is analyzed to verify the correlation.If the data shows strong spatial correlation,space-time series forecasting models can be used to predict the settlement.A test of the settlement of a railway box bridge in Jiangsu province is done to check the method.During the test,six observation monitor spots are installed to monitor the settlement.Results show that the correlation coefficient of the settlement data series from six monitoring spots is as high as 0.99,which means the settlement of the box bridge have a strong spatial trend and correlation.Therefore,space-time series forecasting model is adopted to predict the settlement and comparison is made between the prediction from the model and actual values.Results show that all four error indexes of the space-time prediction are smaller than tradition time series prediction model which proves better reliability and accuracy of the method.
作者
陈倜
Chen Ti(China Railway Design Corporation,Tianjin 300251,China)
出处
《铁道勘察》
2020年第4期37-40,69,共5页
Railway Investigation and Surveying
关键词
铁路构筑物
时空序列
网络模型
沉降预测
运营安全
Railway building
Space-time series
Network model
Deposition forecasting
Safe operation
