Adaptive points range Square-root Cubature Kalman filter for mars approach navigation

The navigation accuracy in approach phase has a significant effect on the successful capture and injection of the Mars spacecraft. On one hand, the perturbation of Mars gravity field on spacecraft's orbit increases when spacecraft gradually approaches the Mars, which results in slow change of the orbit dynamic model. On the other hand, imagines of the Mars and its satellites on sensors vary with the explorer coming to the Mars, which causes change of measurement noise. Since the statistical features of system noise and measurement noise are time-varying, the navigation accuracy will not be optimal if the filter parameters remain constant. In this paper, an adaptive points range Square-root Cubature Kalman filter (AR-SCKF) method is proposed to deal with this problem, which uses several sub-SCKFs with different cubature-points ranges to track the dynamic change of the system noise and measurement noise. The cubature-points ranges of sub-SCKFs can be adjusted automatically according to their navigation performance. Simulations demonstrate that the navigation accuracy of this method has 10% root mean square (RMS) error improvement and 30% maximum error improvement compared to the traditional SCKF. The impact factors of this method like the maximum points range and points range number are also studied.