Comprehensive Prognostic Evaluation of Spacecraft Reaction Wheels Post-Fault Occurrences

This paper investigates the influence of environmental and operational variables on the longevity of spacecraft components, with a specific focus on estimating Remaining Useful Life (RUL), Prognostic Horizon (PH), and State-of-Health (SoH) for reaction wheels afflicted by static friction faults. Emphasizing the significance of fault tolerance in spacecraft hardware, we ascertain the threshold at which component functionality is substantially compromised. Using the Rolling Root-Mean-Square (RRMS) of the reaction wheel’s static friction profile over time as Feature Data (FD) within a 10-minute time frame for two fault scenarios - one-time and repeated occurrences, our analysis demonstrates a direct relationship between fault instances and FD escalation. This correlation is further elucidated by integrating Fault-to-Failure Progression (FFP) FD into the Adaptive Remaining Useful Life Estimator (ARULE) for prognostic assessment, shedding light on the challenges and ramifications of mechanical faults on attitude control systems. ARULE predicts complete degradation in both fault cases within 500 mins of spacecraft flight time. Our study underscores the pivotal role of predictive analytics in spacecraft health management, particularly in optimizing performance and prolonging the service life of critical hardware amidst operational challenges in orbit.