Comparison of algorithms for satellite attitude determination using data from visual sensors

Abstract

The objective of the research was to investigate the efficiency of selected methods of data fusion from visual sensors used on-board satellites for attitude measurements. Data from a sun sensor, an earth sensor, and a star tracker were fused, and selected methods were applied to calculate satellite attitude. First, a direct numerical solution, a numerical and analytical solution of the Wahba problem, and the TRIAD method for attitude calculation were compared used for integrating data produced by a sun sensor and an earth sensor. Next, attitude data from the star tracker and earth/sun sensors were integrated using two methods: weighted average and Kalman filter. All algorithms were coded in the MATLAB environment and tested using simulation models of visual sensors. The results of simulations may be used as an indication for the best data fusion in real satellite systems. The algorithms developed may be extended to incorporate other attitude sensors like inertial and/or GNSS to form a complete satellite attitude system.