Robust two-stage rank filter for Mars entry navigation under parameter uncertainties

Future Mars landing missions require high-precision entry navigation capability, but the uncertainties lie in the Mars atmosphere density and aerodynamic coefficients may cause serious performance degeneration of conventional Kalman filters. This paper proposed a robust two-stage rank filter (RTSRF) out of augmented state rank filter by decoupling the covariance matrices and modifying the measurement update equations. Therefore, this filter could obtain an accurate state estimation by treating the parameter uncertainties as the unknown inputs even the prior knowledge of the inputs are completely unknown, a novel method to dealing with the parameter uncertainties. Then, it was applied for Mars entry navigation simulation, and the results show that the navigation accuracy is greatly improved, fulfilling the requirement of future Mars pinpoint landing missions.