TY - GEN
T1 - Fault Correlation Analysis Algorithm for Multi-Satellite Telemetry Data Based on State Transition Relationships
AU - Qi, Xinhu
AU - Sun, Darui
AU - Chen, Haoyi
AU - Song, Nan
AU - Chen, Yanbin
AU - Zhang, Yanning
N1 - Publisher Copyright:
© 2024 IEEE.
PY - 2024
Y1 - 2024
N2 - In the operational phase of satellites, a substantial amount of telemetry parameters related to physical and state changes are generated, complicating the correlation analysis of faults. This challenge is particularly pronounced in the absence of historical contingency plans, making it difficult to pinpoint relevant telemetry parameters following fault occurrences. This paper presents an algorithm designed for multi-satellite telemetry that conducts correlation analysis of telemetry parameters leading to fault events. The proposed algorithm facilitates multi-satellite telemetry data analysis, establishes strong correlations between fault occurrences and telemetry parameters, and provides interpretable functionalities. Initially, the algorithm conducts state analysis based on actual satellite telemetry data to examine the changes in telemetry parameters. This analysis serves as the foundation for feature encoding, which describes the variations of telemetry parameters within a temporal motion context. Finally, the algorithm employs feature importance analysis within a Random Forest framework to identify telemetry parameters that exhibit a strong correlation with the fault occurrence. This algorithm was applied in a real-world fault scenario in the aerospace domain. The results of telemetry parameters obtained by the algorithm have a strong correlation with the fault occurrence, and have a high consistency with the results inferred by the diagnosis experts.
AB - In the operational phase of satellites, a substantial amount of telemetry parameters related to physical and state changes are generated, complicating the correlation analysis of faults. This challenge is particularly pronounced in the absence of historical contingency plans, making it difficult to pinpoint relevant telemetry parameters following fault occurrences. This paper presents an algorithm designed for multi-satellite telemetry that conducts correlation analysis of telemetry parameters leading to fault events. The proposed algorithm facilitates multi-satellite telemetry data analysis, establishes strong correlations between fault occurrences and telemetry parameters, and provides interpretable functionalities. Initially, the algorithm conducts state analysis based on actual satellite telemetry data to examine the changes in telemetry parameters. This analysis serves as the foundation for feature encoding, which describes the variations of telemetry parameters within a temporal motion context. Finally, the algorithm employs feature importance analysis within a Random Forest framework to identify telemetry parameters that exhibit a strong correlation with the fault occurrence. This algorithm was applied in a real-world fault scenario in the aerospace domain. The results of telemetry parameters obtained by the algorithm have a strong correlation with the fault occurrence, and have a high consistency with the results inferred by the diagnosis experts.
KW - correlation analysis
KW - fault detection
KW - multi-satellite
KW - state transition
KW - telemetry data
UR - http://www.scopus.com/inward/record.url?scp=105000752572&partnerID=8YFLogxK
U2 - 10.1109/ICEACE63551.2024.10898675
DO - 10.1109/ICEACE63551.2024.10898675
M3 - 会议稿件
AN - SCOPUS:105000752572
T3 - 2024 IEEE 2nd International Conference on Electrical, Automation and Computer Engineering, ICEACE 2024
SP - 53
EP - 57
BT - 2024 IEEE 2nd International Conference on Electrical, Automation and Computer Engineering, ICEACE 2024
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2nd IEEE International Conference on Electrical, Automation and Computer Engineering, ICEACE 2024
Y2 - 29 December 2024 through 31 December 2024
ER -