ADAPTIVE SENSOR FUSION FOR ATTITUDE ESTIMATION OF SMALL SATELLITES
small satellites; spacecraft dynamics; attitude estimation; Kalman filter; sensor fusion; adaptive Kalman filter.
Due to the technology progress, particularly in electronics the satellite development has become more accessible, allowing its manufacturing by startup companies and universities. Such satellites use low cost components, for general purpose, where one seeks for exchanging a high cost and high reliability platform for a low cost vehicle constellation, less reliable, more advanced technologically and with easy replacement. In this context, an adaptive attitude determination system is an alternative deal with sensors that have time-varying statistics or they are not completely known. An adaptive ADCS grants more reliability to the platform, increasing its life-cycle and reducing the replacement cost. Taking it as motivation, the work presents the analysis of algorithms for adaptation of the filtering method in sensor measurements in the attitude estimation of a small satellite. It is limited to the study of stochastic methods for such goal, as the estimation of the process and observation noise covariance matrices, along with the hypothesis test regarding the fault occurrence and the matrices adaptation by multiple scale factors. The results obtained show that the covariance matrices adaptation allows the faulty measurements to be accommodated, reducing their influence in the filter and providing the attitude accuracy preservation for the satellite when comparing with the estimator without the fault detection and isolation mechanism.