Deep Clustering-Based Anomaly Detection and Health Monitoring for Satellite Telemetry

Satellite telemetry data plays an ever-important role in both the safety and reliability of a satellite. These two factors are extremely significant field space systems missions. Since it is challenging to repair orbit, health monitoring early anomaly detection approaches crucial for success A large number efficient accurate methods have been proposed aerospace but without showing enough concern patterns that can be mined from normal operational data. Concerning this, present paper proposes DCLOP, intelligent Deep Clustering-based Local Outlier Probabilities approach aims at detecting anomalies alongside extracting realistic reasonable The combines (i) new deep clustering method uses dynamically weighted loss function with (ii) adapted version based on results clustering. DCLOP effectively monitors status spacecraft detects warnings its on-orbit failures. Therefore, this enhances validity accuracy systems. performance suggested assessed using actual cube satellite experimental findings prove competitive currently used techniques terms effectiveness, viability, validity.