Robust fault tolerance alignment for INS with sensor faults

In this paper, a robust fault tolerance initial alignment approach is proposed for a class of inertial navigation systems (INSs) with multiple disturbances and sensor faults. Based on modeling and noise analysis, the drifts, unmodeled dynamics, parametric uncertainties, external disturbances are formulated into different types of disturbances described by exo-system, stochastic and norm-bounded variables, respectively. In the proposed fault tolerance alignment approach, the drift and fault observers are designed to estimate the inertial sensor drift and fault for compensating and accommodating, respectively. The robust filter is designed to optimize the estimation error and attenuate norm bounded uncertain disturbances. Finally, simulations and comparisons with unscented Kalman filter (UKF) for ground stationary base alignment of an INS are provided to show the efficiency of the proposed approach.