GNSS观测中移动平均法和小波变换的滤噪效果比较

Comparison of Filtering Effect between Moving Average Method and Wavelet Transform in GNSS Observation

多路径误差是GNSS变形监测中的主要误差源。针对GNSS观测中多路径误差的特性,该文采用了移动平均法和小波变换法去除观测噪声,提取多路径效应用于改正去噪后的历元坐标。实验结果表明:移动平均法和小波变换都能很好地去除坐标域内的观测噪声,通过两种方法去噪前后的对比,得出小波变换去除观测噪声的效果明显优于移动平均法。

The multipath error is the main source of error in GNSS deformation monitoring. Amied to the characteristics of multipath error in GNSS observation,this paper adopts the moving average method and wavelet transform to remove the observation noise,and the multipath effect is used to correct the epoch coordinates after denoising. The experimental results show that the moving average method and wavelet transform all can remove the observed noise in the coordinate domain well. Through the comparison of the two methods before and after denoising,it concludes that the effect of wavelet transformation on removing the observation noise is obviously better than the moving average method.