Nonlinear attitude estimation algorithm for gyroless satellite and its simulation

Aim. The severe nonlinearities of gyroless satellite attitude model, which degrade the convergence and accuracy of EKF (extended Kalman filter) and UKF (unscented Kalman filter) respectively, and the large initial state errors sometimes cause estimation failure. We adopt a new kind of nonlinear estimation algorithm to deal with the problem. Section 2 of the full paper explains our new estimation algorithm. Subsection 2.2 solves a nonlinear smoothing problem for the current and previous sample intervals using iterative numerical techniques and retains the nonlinearities of a number of stages that precede the stage of interest without any approximation. Subsection 2.3 accelerates computing speed through analyzing the influencing factors. The results of simulation, given in Figs. 1 through 3, illustrate preliminarily that the new algorithm exhibits good convergence and accuracy for gyroless satellite attitude estimation problems that have large initial state errors and severe nonlinearities of satellite's dynamics model.