An Optimization Scheme of the Loosely Coupled SINS/DVL Integrated Navigation Based on ST-EKF (September 2023)

Aiming at the state transformation extended Kalman filter (ST-EKF)-based SINS/DVL loosely integrated navigation algorithm, the construction of the measurement equations faces the problem of model inaccuracy, thereby reducing the final navigation accuracy. Therefore, an optimization scheme for the loosely coupled SINS/DVL integrated navigation based on ST-EKF is proposed. Its matching mode differs from the existing loosely coupled SINS/DVL integrated navigation based on ST-EKF. It adopts a matching mode based on the actual navigation coordinate frame's velocity information. When DVL is in different working modes, measurement equations based on the navigation coordinate frame's velocity error are established, improving the model accuracy of measurement equations under different working modes. Experimental results show that the proposed optimized scheme can improve the positioning accuracy by 65.8% compared with the EKF-based SINS/DVL integrated navigation system, and by 61.7% compared with the existing ST-EKF-based SINS/DVL integrated navigation system.