基于EWN模型的高铁沉降观测数据预测研究

The Research on High Speed Railway Settlement Observation Data Prediction based on EWN Model

针对高铁沉降变形监测数据存在的误差将降低预测模型预测精度的问题,利用EMD算法结合小波去噪算法对高铁累积沉降数据进行去噪预处理,利用NAR预测模型完成去噪后数据的预测实验。实验结果表明,EWN模型相对仅使用小波去噪的NAR模型具有更高的预测水平,其中平均相对误差减少了4.6%,残差均值减少了0.29mm。将EMD算法结合小波去噪算法应用于NAR模型的去噪预处理,可以提高NAR模型的预测精度。

Aiming at the problem that the error of high-speed railway settlement deformation monitoring data will reduce the prediction accuracy of prediction model,EMD algorithm and wavelet denoising algorithm are used to denoise the high-speed railway cumulative settlement data,and NAR prediction model is used to complete the prediction experiment of denoised data.The experimental results show that the ewn model has a higher prediction level than the NAR model using only wavelet denoising,in which the average relative error is reduced by 4.6%and the mean residual error is reduced by 0.29mm.The EMD algorithm combined with wavelet denoising algorithm is applied to the denoising preprocessing of NAR model,which can improve the prediction accuracy of NAR model.