Bias Compensated Inertial Navigation for Venus Balloon Missions

In this paper we investigate various strategies for reducing navigation errors for a future Venus balloon mission equipped with a low-mass MEMS IMU (e.g. STIM300), and focus on two of these mitigation strategies in detail. First, we propose to mechanically rotate the IMU while collecting data thereby averaging out the bias contributions and increasing accuracy. To this end, we experimentally provide proof-of-concept by rotating a STIM300 IMU about a single-axis and show that navigation performance for this idealized scenario is substantially improved. Second, we propose to use on-board accelerometer measurements to exploit the projection of the gravity field along the axes of the IMU frame to provide (2 axis) attitude information. To this end, we designed a Kalman filter for a hovering balloon scenario where the accelerometer measurements are modeled as explicit functions of attitude and are incorporated as direct measurements in the filter and show that this strategy leads to improved navigation performance. The results given here are part of a longer-term effort to understand and improve navigation performance on a Venus balloon and should be considered as first steps toward this goal.