INS Drift Mitigation During DVL Outages

An inertial navigation system is commonly fused with a Doppler velocity log to provide an accurate navigation solution for autonomous underwater vehicles. In some situations, like when passing or over trenches on the seafloor, not all of the Doppler velocity log beam measurement are reflected back. As a consequence, the navigation solution depend only on the inertial navigation solution, which drifts in time due to measurement errors in the inertial sensors. To cope with such situations, an algorithm to enable the estimation of the velocity vector in situations of complete outage was proposed, based on past Doppler velocity log measurements and a kinematic model. In this paper, this algorithm is elaborated by taking past filter based estimated velocity measurements. In this manner, the inertial data is incorporated into the past velocity measurements and thus richer information is obtained. However, in such situations the process and measurement noises are now correlated and hence, should be accounted for in the navigation filter. Simulation results are presented to show the benefits of using the proposed approach.