2D Estimation of Velocity Relative to Water and Tidal Currents Based on Differential Pressure for Autonomous Underwater Vehicles | IEEE Journals & Magazine | IEEE Xplore

2D Estimation of Velocity Relative to Water and Tidal Currents Based on Differential Pressure for Autonomous Underwater Vehicles


Abstract:

Reliable navigation of autonomous underwater vehicles (AUVs) depends on the quality of their state estimation. Providing robust velocity estimation thus plays an importan...Show More

Abstract:

Reliable navigation of autonomous underwater vehicles (AUVs) depends on the quality of their state estimation. Providing robust velocity estimation thus plays an important role. While water currents are main contributors to the navigational uncertainty of AUVs, they are also an important variable for oceanographic research. For both reasons, water current estimation is desirable during AUV operations. State of the art velocity estimation relies on expensive acoustic sensors with considerable energy requirements and a large form factor such as Doppler Velocity Logs (DVL) and Acoustic Doppler Current Profilers (ADCP), while water currents are either estimated with the same sensors, or with algorithms that require accurate position feedback. In this letter, we introduce a low-cost, lightweight and energy efficient sensor (DPSSv2) to estimate fluid relative velocity in 2D based on differential pressure. The sensor is validated in field trials on-board an AUV in the presence of tidal currents. We further show that, while moving against the currents, our device is capable of estimating tidal currents in situ with comparable accuracy to a DVL, given a source for absolute vehicle velocity. Additionally, we establish the limitations of the current design of DPSSv2 while moving with the currents.
Published in: IEEE Robotics and Automation Letters ( Volume: 5, Issue: 2, April 2020)
Page(s): 3444 - 3451
Date of Publication: 26 February 2020

ISSN Information:

Funding Agency:


References

References is not available for this document.