DVL-aided Navigation Filter for Maritime Applications ?

摘要

This publication presents the ongoing developments of a tightly-coupled navigation filter for maritime applications. The filter is based on quantities of an Inertial Measurement Unit (IMU) and is aided by signals of a Global Navigation Satellite System (GNSS). Furthermore, to improve robustness, the filter is aided by data of a Doppler Velocity Log (DVL), which determines vessel velocity in three dimensions relative to ground. This publication introduces the filter algorithms and the necessary extensions to include DVL information. Therefore, the received data are synchronized, the measurement noise is parameterized for use in the applied Extended Kalman Filter (EKF) and the improved filter-convergence behavior at system initialization is investigated and proofed. The new filter approach is evaluated and compared within extensive real-world measurement campaigns, which are carried out in the Harbor of Rostock, Germany. Results show an improved convergence behavior, a more robust localization in case of poor satellite reception due to a more accurate velocity correction, and slightly improved position estimation accuracy, depended on the GNSS observables accuracy within the navigation filter.