scholarly journals On-line identification of autonomous underwater vehicles through global derivative-free optimization

2013 ◽  
Author(s):  
George C. Karras ◽  
Charalampos P. Bechlioulis ◽  
Matteo Leonetti ◽  
Narcais Palomeras ◽  
Petar Kormushev ◽  
...  
Keyword(s):  
Autonomous Underwater Vehicles ◽  
Underwater Vehicles ◽  
Line Identification ◽  
Derivative Free Optimization ◽  
Derivative Free ◽  
On Line

scholarly journals Combining Local and Global Direct Derivative-Free Optimization for Reinforcement Learning

2012 ◽  
Vol 12 (3) ◽  
pp. 53-65 ◽  
Author(s):  
Matteo Leonetti ◽  
Petar Kormushev ◽  
Simone Sagratella
Keyword(s):  
Reinforcement Learning ◽  
Autonomous Underwater Vehicle ◽  
Underwater Vehicle ◽  
Target Area ◽  
Policy Space ◽  
Derivative Free Optimization ◽  
Derivative Free ◽  
Actual Environment ◽  
On Line ◽  
On Line Learning

Abstract We consider the problem of optimization in policy space for reinforcement learning. While a plethora of methods have been applied to this problem, only a narrow category of them proved feasible in robotics. We consider the peculiar characteristics of reinforcement learning in robotics, and devise a combination of two algorithms from the literature of derivative-free optimization. The proposed combination is well suited for robotics, as it involves both off-line learning in simulation and on-line learning in the real environment. We demonstrate our approach on a real-world task, where an Autonomous Underwater Vehicle has to survey a target area under potentially unknown environment conditions. We start from a given controller, which can perform the task under foreseeable conditions, and make it adaptive to the actual environment.

Three-dimensional leader–follower formation control of multiple autonomous underwater vehicles based on line-of-sight measurements using the backstepping method

2018 ◽  
Vol 232 (7) ◽  
pp. 819-829 ◽  
Author(s):  
Shikun Pang ◽  
Jian Wang ◽  
Jingyang Liu ◽  
Hong Yi
Keyword(s):  
Formation Control ◽  
Autonomous Underwater Vehicle ◽  
Autonomous Underwater Vehicles ◽  
Three Dimensional ◽  
Underwater Vehicles ◽  
Line Of Sight ◽  
Backstepping Method ◽  
Virtual Reference ◽  
On Line ◽  
Backstepping Controller

This article presents the three-dimensional leader–follower formation control strategy of multiple autonomous underwater vehicles on the basis of line-of-sight measurement. Sensors on autonomous underwater vehicles can measure line-of-sight range and bearing information, and autonomous underwater vehicles in formation do not require mutual exchange with one another, which could avoid the problem of data transmission delay. The proposed Lyapunov-based backstepping approach is developed for the formation controller. The backstepping controller propels the follower autonomous underwater vehicles to follow the desired virtual reference route provided according to the range and bearing information of the leader autonomous underwater vehicle and the predetermined formation. Simulation results indicate that follower autonomous underwater vehicles can travel along the desired route and successfully achieve the designed formation shape. The effectiveness of the employed method is validated.

ON-LINE IDENTIFICATION OF HYDRODYNAMICS IN UNDERWATER VEHICLES

2005 ◽  
Vol 38 (1) ◽  
pp. 344-349 ◽  
Author(s):  
Mario A. Jordán ◽  
Jorge L. Bustamante ◽  
Edwin Kreuzer ◽  
Volker Schlegel
Keyword(s):  
Underwater Vehicles ◽  
Line Identification ◽  
On Line
Sign in / Sign up

Export Citation Format

Share Document