.png){kind=icon}
.png){kind=icon}
Return
Adaptive filtering method for SLR data based on local density
.png){kind=icon}
PRE
AI
DOI：10.1016/j.optcom.2025.132735.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Conventional satellite laser ranging (SLR) data preprocessing typically uses manual screen processing methods. However, this method has a low degree of automation, limited processing efficiency, and is easily influenced by the subjective experience of researchers. To address the above issues, this paper proposes an adaptive filtering method for SLR data based on local density. This method is based on the clustered distribution characteristics of valid echo signals from SLR and introduces a density comparison method to improve the traditional DBSCAN algorithm. By simulating the valid echo points of SLR data segments, it automatically calculates the density comparison threshold and uses two rounds of filtering to remove noise signals. The experimental results show that for observation data from satellites in different orbits, this method achieves a recognition rate of over 85 % and a misdetection rate of no more than 4 %, with processing results superior to manual screen processing, the traditional DBSCAN algorithm, and the Graz algorithm. The internal precision of normal points is improved by up to 44.91 % compared to the results from the 7237 station, by up to 37.34 % compared to the traditional DBSCAN algorithm, and by up to 23.48 % compared to the Graz algorithm. In addition, ablation studies and sensitivity analyses have also validated the necessity of the method's two-stage architecture and the selection of its key parameters. Furthermore, the method can batch-process SLR data with an average processing efficiency of 26.7 s per data file, meeting the requirements for efficient automated SLR processing. This method provides an efficient solution for real-time data processing in future high-frequency SLR systems.
Keywords:
Satellite laser ranging
Data processing
Adaptive filtering
Journal
IF:
2.5
Papers:
595
Citations:
2.7W