Robust Control of Autonomous Remotely Operated Vehicles at Exposed Aquaculture Sites.
Résumé
This paper presents two adaptive nonlinear controllers for robust velocity and heading control of a remotely operated vehicle navigating in exposed aquaculture sites. The controllers are designed to make it possible for the vehicle to execute autonomous traversal of an aquaculture net pen using a net-following guidance algorithm, in the presence of significant environmental disturbances. Each of these controllers is tailored for different physical models depending on the application scenario. The first is suitable for slow-speed maneuvers and the second applies to cases in which aggressive maneuvers are needed. The second model has the added difficulty that additional Coriolis-forces nonlinearities must be considered. Stability proofs for the closed-loop system under the action of each controller are provided. For the first controller, the closed-loop system is proven to be uniformly globally asymptotically stable and uniformly locally exponentially stable at the origin. For the second controller the origin of the closed-loop system is proven to be uniformly globally stable and asymptotic convergence of the velocity and heading error states are provided for the second control law. In addition, the paper presents simulation and experimental results to validate and illustrate the theoretical analysis, where the controllers are applied to an industrial underwater robot. In particular, the first control law was successfully tested at a full-scale aquaculture site under realistic operational conditions.
Origine | Fichiers produits par l'(les) auteur(s) |
---|