Robust initial alignment for inertial navigation systems with multiple types of disturbances

In this paper, the robust initial alignment problem for inertial navigation systems (INSs) is investigated. Different from most previous works only focusing on Gaussian noises (or other single disturbance), the INS error equation with multiple types of disturbances is constructed. We consider two types of disturbances, where the first type is the sensor drifts described by an exosystem, and the second one includes the nonlinear modeling uncertainties and other measurement noises. A robust filter is constructed for the concerned INS with disturbance rejection and attenuation performance. In the proposed approach, the drift estimations are applied to reject the inertial sensor drifts, H_∞ performance is applied to attenuate the norm bounded uncertain disturbances. The proposed multi-objective filter is with both disturbance rejection and attenuation performance. Finally, simulation for stationary base alignment of an INS is given to show the efficiency of the proposed approach.