Abstract
Underwater gliders have proven their outstanding merit as efficient long-endurance ocean sampling vehicles. Yet, their limited payload and energy constraints reduce capabilities for accurate path following. For glider operations in narrow environments such as long fjords, there is a stricter requirement on the ocean sampling path. This paper provides a path-following solution for gliders with limited navigational payloads. The proposed solution unfolds in two parts: (i) a recurrent neural network is presented to enhance the dead-reckoning performance of the glider and alleviate the guidance problems related to the lack of navigation sensors; (ii) an integral line-of-sight path-following controller is adopted to compensate drift due to unknown ocean currents. Simulation studies are carried out to evaluate the performance of the proposed system.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call