Abstract | ||
---|---|---|
This paper proposes a field application of a high-level Reinforcement Learning (RL) control system for solving the action selection problem of an autonomous robot in a cable tracking task. The underwater vehicle ICTINEUAUV learns to perform a visual based cable tracking task in a two step learning process. First, a policy is computed by means of simulation where a hydrodynamic model of the vehicle simulates the cable following task. Once the simulated results are accurate enough, in a second step, the learned-in-simulation policy is transferred to the vehicle where the learning procedure continues in a real environment, improving the initial policy. The natural actor-critic (NAC) algorithm has been selected to solve the problem in both steps. This algorithm aims to take advantage of policy gradient and value function techniques for fast convergence. Actor's policy gradient gives convergence guarantees under function approximation and partial observability while critic's value function reduces variance of the estimates update improving the convergence process. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1109/ROBOT.2010.5509751 | Robotics and Automation |
Keywords | Field | DocType |
convergence of numerical methods,function approximation,learning (artificial intelligence),mobile robots,remotely operated vehicles,underwater vehicles,AUV,ICTINEU,action selection problem,autonomous robot,convergence process,critic value function,function approximation,high level reinforcement learning control system,learned- in-simulation policy,natural actor critic learning,underwater cable tracking,underwater vehicle,vehicle hydrodynamic model | Convergence (routing),Remotely operated underwater vehicle,Observability,Function approximation,Control theory,Control engineering,Engineering,Robot,Autonomous robot,Mobile robot,Reinforcement learning | Conference |
Volume | Issue | ISSN |
2010 | 1 | 1050-4729 E-ISBN : 978-1-4244-5040-4 |
ISBN | Citations | PageRank |
978-1-4244-5040-4 | 1 | 0.41 |
References | Authors | |
11 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Andres El-fakdi | 1 | 33 | 5.59 |
Marc Carreras | 2 | 374 | 31.66 |
Enric Galceran | 3 | 236 | 13.50 |