Fault-tolerant ADCS for Small Satellites Using Bayesian Networks

The Attitude Determination and Control System (ADCS) in satellites uses different modes to estimate and control attitude during various mission phases. This study presents an ADCS design for an Earth-observing small satellite, including different modes with controllers and estimators. Mode transitions are achieved using threshold values and a discrete state Markov Chain model, with Bellman optimization controlling the Markov chain. Additionally, a Fault Detection and Isolation (FDI) algorithm is proposed, utilizing a Bayesian Network (BN) constructed with expert knowledge. A simulation environment is created to evaluate the performance of the algorithms and modes.