GPS结构振动监测数据滤波方法及其性能实验研究

Abstract

使用GPS监测结构振动的目的在于提取振动信号特征,但GPS观测量受多种误差源的影响,因此,选用合理的数据处理方法有效地分离各误差项,对于提高GPS的监测精度具有重要意义。将Vondrak滤波、小波滤波、自适应FIR滤波和卡尔曼滤波等四种方法应用于资料序列中振动信号的分离,通过对模拟振动实验观测资料的分析表明:运用滤波法可提高GPS测量微小动态变形和变频振动信号的检测能力;4种滤波法均能有效地提高GPS监测结构振动的精度,其中,Vondrak滤波和小波滤波的性能相当,且优于自适应FIR和卡尔曼滤波。同时,在对各滤波法参数选择的优缺点进行分析的基础上,提出不同情况下选择滤波器的建议。 The purpose of GPS structural vibration monitoring is to obtain information on the frequency and amplitude of vibrations based on GPS observations that are often affected by various errors.Filters are frequently used to improve GPS accuracy and to retrieve vibration signals from GPS observational series.This paper studies the performances of four commonly used filters,i.e.,Vondrak,wavelet,adaptive FIR and Kalman filters,for such applications.Controlled experiments are carried out and the results show that the capability of GPS in tracking structural dynamics and complex signals can be improved with any of the filters.The performances of Vondrak and wavelet filters are almost the same and superior to the adaptive FIR and Kalman filters.Recommendations are given for the selection of filters and filter parameters for different situations based on an analysis of the advantages and disadvantages of each of the filters.