小波-时序组合模型在地铁变形监测预测中的应用研究

Application of Wavelet-timing Combination Model to Prediction of Metro Deformation Monitoring

以郑州地铁2号线实际观测数据为基础,用小波分析的方法对其做降噪及剔除粗差的处理,分析不同小波基函数、分解层次、阈值选取方法对小波去噪结果的影响,选取最佳的小波模型[1]。分别用实测数据和去噪后的数据用时间序列模型做出预测,并与实测数据做对比,结果表明:小波去噪后的时间序列模型预测精度更高,说明该组合模型适用于地铁沉降变形监测的预测分析。

Zhengzhou metro line 2 to the observed data as the foundation, using the wavelet analysis method for the noise reduction and elimination of gross error handling, analysis, decomposition levels, different wavelet basis function method for wavelet denoising threshold value results, the influence of selecting the best wavelet model [1] . Respectively with the measured data and data after denoising with time series model to make predictions, and comparing with experimental data and the results show that after wavelet denoising time series prediction model accuracy is higher, shows that the combined model prediction analysis of the subway settlement deformation monitoring has certain applicability.