基于小波变换的变形监测数据误差处理方法研究

Research on Error Processing Method of Deformation Monitoring Data Based on Wavelet Transform

基于北斗卫星导航系统的高边坡在线监测大大提高了监测效率,但由于卫星信号遮挡以及其他因素干扰问题,原始监测数据中容易存在高频噪声、粗差和周期性误差等,从而降低监测数据质量。本文利用小波变换对原始观测数据进行分解,采用低频层对监测数据进行恢复,剔除粗差影响;然后对剔除粗差后的数据进行二次小波变换,采用不同的滤波算法对变换后的各层系数分别进行滤波处理,将处理后的数据与原始数据进行对比分析,验证了两次小波变换误差处理后的监测数据曲线更加平缓,周期性误差显著降低,在保留原有变形特征的情况下大大提升了数据质量,为高边坡安全态势的分析提供更为直观可靠的监测数据。

The online monitoring of high slopes based on the Beidou satellite navigation system has greatly improved the monitoring efficiency.However,due to the occlusion of satellite signals and the interference of other factors,high-frequency noise,gross errors and periodic errors are likely to exist in the original monitoring data,which reduces the monitoring data quality.In this paper,wavelet transform is used to decompose the original observation data,and the low-frequency layer is used to restore the monitoring data,and the influence of gross errors is eliminated;then the data after eliminating gross errors is subjected to secondary wavelet transform,and different filtering algorithms are used to transform each layer.The coefficients are filtered separately,and the filtered data and the original data are compared and analyzed,and it is verified that the monitoring data curve after two wavelet transform error processing is smoother,the periodic error is significantly reduced,and the original deformation characteristics are retained.The data quality is improved,and more intuitive and reliable monitoring data is provided for the analysis of the safety situation of high slopes.