Square Root Spherical Simplex Unscented Kalman Filter for Micro Satellite Attitude Measurement

A square root spherical simplex unscented Kalman filtering for micro satellite attitude determination was presented. The filter performance was guaranteed and the computation was reduced by means of spherical simplex unscented transformation (SSUT) and modified Rodrigues parameters (MRPs). The square root technique improved the filter´s numerical stability by guaranteeing positive-semi definiteness of the state covariance. The simulation results of some micro-satellite showed that this algorithm could insure accuracy, fast convergence and high robustness with high computation efficiency, which was suitable for the attitude estimation of micro-satellite.