Calibration of Rotation MEMS-IMU/DVL System Based on FGO

Abstract

Traditional underwater navigation relies on a costly fiber optic gyroscope (FOG) inertial navigation system (INS)/Doppler velocity log (DVL) system. This study aims to address the challenges of applying rotation INS technology within MEMS-IMU/DVL systems, potentially enabling them to replace FOG systems—something that nonrotating MEMS-IMUs cannot achieve. While integrating rotation INS techniques enhances accuracy, it also introduces system errors that affect precision and are difficult to calibrate in field swaying conditions. To facilitate calibration in these situations, we incorporate an external nonrotating MEMS-IMU alongside the rotating MEMS-IMU, forming a dual-IMU system for more reliable calibration. We also design a specialized factor graph optimization structure that estimates the INS state using maximum a posteriori estimation and employs maximum likelihood estimation for constant calibration parameters, effectively minimizing the impact of inaccurate prior information in standard estimation methods. After calibration and compensation, the compact MEMS-IMU/DVL system achieves high positional accuracy and delivers smoother heading results.