A Sliding Window Optimal Tracking Differentiator Filtering Method for Satellite Telemetry Data

Ke Zhang,Haixu Jiang,Jingyu Wang

doi:10.1051/jnwpu/20203830515

Abstract

The initial satellite telemetry data acquired by ground stations usually contain noise and outlier interference. In order to ensure the accurate analysis of satellite status, the telemetry data need to be filtered. In this paper, a sliding window optimal tracking differentiator filtering (SWOTDF) method for satellite telemetry data is proposed. Aiming at the problem of parameter selection during the filtering of the optimal tracking differentiator, the amplitude-frequency characteristics of the maximum tracking differentiator are analyzed by sine sweep frequency method, and the mapping relationship between tracking factors and filtering effects is established. On this basis, the telemetry data are divided by sliding windows, and the relationship between local stability of data in each window and tracking factors is further analyzed. The calculation method of local data tracking factor is given to realize dynamic optimal tracking differentiator filtering of telemetry data in each window. Experimental results show that the SWOTDF method can effectively avoid the limitations of traditional digital filters in processing nonlinear telemetry data, and can effectively filter out noise and outliers in satellite telemetry data.

Highlights

The initial satellite telemetry data acquired by ground stations usually contain noise and outlier interfer⁃ ence
In order to ensure the accurate analysis of satellite status, the telemetry data need to be filtered
Aiming at the problem of parameter selection during the filtering of the optimal tracking differentiator, the amplitude⁃frequency characteristics of the maximum tracking differentiator are analyzed by sine sweep frequency method, and the mapping relationship between tracking factors and filtering effects is established

Summary

Introduction

最速跟踪微分器[10⁃11] 是由韩京清提出的一种二阶离散形式跟踪微分器,具有良好的数据跟踪能力,通过选用合理的参数,最速跟踪微分器能够实现对含噪数据的理想滤波和去野。但目前最速跟踪微分器的参数选择仍以经验法为主,并且在滤波时采用全局固定参数,这都直接影响最速跟踪微分器的滤波效果,使最速跟踪微分器的应用范围受到局限, 本文提出一种基于滑动最速跟踪微分器的遥测数据滤波方法,通过滑动窗口对数据进行划分,根据各窗口内数据的稳定性设计局域最速跟踪微分器的参数,实现对卫星遥测数据的理想滤波。图 6 y1 容错 Q⁃滤波结果图 7 y1 滑动最速跟踪微分器滤波结果时,令周期 T = 1 250,周期数量 K ∈ [0,1,2,3,4] , 得到矩形仿真数据 y2(ti),如图 8 所示。对仿真数据 y2(ti) 进行滤波时, 同样设定各方法的滑动窗宽度为 50, 滑动步长为 1, 得到滤波结果如图 9 至 12 [2] DING F, WANG F, XU L, et al Decomposition Based Least Squares Iterative Identification Algorithm for Multivariate Pseudo⁃ Linear Arma Systems Using the Data Filtering[ J] .

Results

Conclusion