利用小波变换进行GPS振动观测值的异常探测

本文利用小波变换技术进行GPS振动观测值的异常探测,将GPS振动观测值序列视为一组观测信号,先利用小波低频分解提取信号趋势,再利用小波高频分解探测异常值的位置,提出结合观测体自身变形趋势判断异常值是否存在,再通过小波高频分解和小波阈值去噪找到异常值准确位置的探测方法。结合某GPS振动平台上的振动观测实验数据进行仿真分析,获得准确的异常探测结果。实践证明文中研究的异常值探测方法是可行的和有效的,为GPS振动观测值的异常探测提供新思路。

Using wavelet transform to detect GPS vibratory observation data

This paper presents the wavelet transform technology to detect GPS vibratory observation data , regarding GPS vibratory observation data order as a group of observation signal .Firstly it studies how to extract the signal trend by wavelet decomposition of low frequency ,and then use high frequency wavelet decomposition to detect the location of abnormal data .A method of detection abnormal data combined with the construction deformation trend is proposed to judge w hether the abnormal data exist ,and then to find abnormal data accurately position through high frequency wavelet decomposition and wavelet threshold denoising .Combining the vibratory observation experimental data in some GPS vibration platform , it mates a simulation analysis to obtain an accurate anomaly detection result . T his method proves to be effective and feasible ,w hich can provide a reference for the anomaly detection ..