No abstract

No abstract

No abstract

Inertial navigation system (INS) is the most common positioning system in the underwater vehicle due to the limitations of penetration of the global navigation system. Amendment of INS cumulative error is made possible using data of auxiliary sensors such as Doppler velocity log (DVL). Due to the environmental conditions of the seabed, access to full DVL beams is impractical. Moreover, the statistical characteristics of measuring noise are unknown. These dramatically reduce the performance of the navigation system. Therefore, a loosely coupled (LC) approach is used to fuse partial DVL measurements in this study. This approach benefits from a virtual beam containing the velocity obtained in the last INS step instead of inaccessible DVL beams. To deal with the unknown statistical characteristics of the measurement and the limitations of access to DVL beams, an adaptive robust Kalman filter (ARKF) based on the Huber cost function is proposed. The performance of the proposed integrated navigation system for a remotely operated vehicle is investigated in the presence of observation noise and outliers. Results show that the proposed ARKF is robust against vigorous maneuvers, improves the estimation accuracy effectively, and can effectively reject measurement outliers.

No abstract