Borja-videos - Historial de revisiones

Manuel.grana: /* B) Linked Robots */

2012-01-22T00:36:12Z

B) Linked Robots

← Revisión anterior		Revisión del 02:36 22 ene 2012
Línea 84:		Línea 84:

	*B.2 Max. tangential speed for last robot was limited (50%). References move full-speed. ([[media:2010.5.run5.avi]])		*B.2 Max. tangential speed for last robot was limited (50%). References move full-speed. ([[media:2010.5.run5.avi]])
	The last robot is forced to move slower than the rest and, because of this, the robots aren't capable of catching the references. Error ~~spreads among~~ the system.		The last robot is forced to move slower than the rest and, because of this, the robots aren't capable of catching the references. Error propagates across the system.

	*B.3 Max. tangential speed for last robot was limited (50%). References move at 75% speed.([[media:2010.5.run6.avi]])		*B.3 Max. tangential speed for last robot was limited (50%). References move at 75% speed.([[media:2010.5.run6.avi]])

Manuel.grana: /* Consensus-based approach to L-MCRS systems */

2012-01-22T00:34:45Z

Consensus-based approach to L-MCRS systems

← Revisión anterior		Revisión del 02:34 22 ene 2012
Línea 73:		Línea 73:
	=Consensus-based approach to L-MCRS systems=		=Consensus-based approach to L-MCRS systems=

	These are some examples of real life experiences on the hose transportation problem. Robot detection and control software is run on a PC. Red dots represent the references (where robots "should be") and green dots the ~~posture given by~~ the camera (where "they are"). Commands are sent to robots using radio transceivers:		These are some examples of real life experiences on the hose transportation problem. Robot detection and control software is run on a PC. Red dots represent the references (where robots "should be") and green dots the position computed from the information obtained from the overhead camera (where "they are"). Commands are sent to robots using radio transceivers:
	==A) Non-Linked Robots==		==A) Non-Linked Robots==

Bortx en 13:32 2 ene 2012

2012-01-02T13:32:19Z

← Revisión anterior		Revisión del 15:32 2 ene 2012
Línea 5:		Línea 5:
	==Partially Constrained Models (2011)==		==Partially Constrained Models (2011)==

	In these videos, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task. ~~This video is~~ intended as complementary material for the paper submitted to the Robotics and Autonomous Systems journal		In these videos, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task. These videos are intended as complementary material for the paper submitted to the Robotics and Autonomous Systems journal

	Transfer Learning with Partially Constrained Models: application to reinforcement learning of linked multicomponent robot system control		Transfer Learning with Partially Constrained Models: application to reinforcement learning of linked multicomponent robot system control
	Borja Fernandez-Gauna, Jose Manuel Lopez-Guede, Manuel Graña		Borja Fernandez-Gauna, Jose Manuel Lopez-Guede, Manuel Graña

	-First trial-goal example. Video: [[media:~~PCM~~-~~RAS~~.avi]]		-First trial-goal example. Video: [[media:RAS_-_PCM_1c.avi]]

	-Second trial-goal example. Video: [[media:RAS_-_PCM_2c.avi]]		-Second trial-goal example. Video: [[media:RAS_-_PCM_2c.avi]]

Bortx: /* Partially Constrained Models (2011) */

2012-01-02T12:50:33Z

Partially Constrained Models (2011)

← Revisión anterior		Revisión del 14:50 2 ene 2012
Línea 5:		Línea 5:
	==Partially Constrained Models (2011)==		==Partially Constrained Models (2011)==

	In ~~this video~~, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task. This video is intended as complementary material for the paper submitted to the Robotics and Autonomous Systems journal		In these videos, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task. This video is intended as complementary material for the paper submitted to the Robotics and Autonomous Systems journal

	Transfer Learning with Partially Constrained Models: application to reinforcement learning of linked multicomponent robot system control		Transfer Learning with Partially Constrained Models: application to reinforcement learning of linked multicomponent robot system control
	Borja Fernandez-Gauna, Jose Manuel Lopez-Guede, Manuel Graña		Borja Fernandez-Gauna, Jose Manuel Lopez-Guede, Manuel Graña

	Video: [[media:PCM-RAS.avi]]		-First trial-goal example. Video: [[media:PCM-RAS.avi]]

			-Second trial-goal example. Video: [[media:RAS_-_PCM_2c.avi]]

	==Round-Robin Distributed Q-Learning (2011)==		==Round-Robin Distributed Q-Learning (2011)==

Manuel.grana: /* Partially Constrained Models (2011) */

2011-12-30T19:58:40Z

Partially Constrained Models (2011)

← Revisión anterior		Revisión del 21:58 30 dic 2011
Línea 5:		Línea 5:
	==Partially Constrained Models (2011)==		==Partially Constrained Models (2011)==

	In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task.		In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task. This video is intended as complementary material for the paper submitted to the Robotics and Autonomous Systems journal

			Transfer Learning with Partially Constrained Models: application to reinforcement learning of linked multicomponent robot system control
			Borja Fernandez-Gauna, Jose Manuel Lopez-Guede, Manuel Graña

	Video: [[media:PCM-RAS.avi]]		Video: [[media:PCM-RAS.avi]]

Bortx: /* Partially Constrained Models (2011) */

2011-12-30T11:13:20Z

Partially Constrained Models (2011)

← Revisión anterior		Revisión del 13:13 30 dic 2011
Línea 7:		Línea 7:
	In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task.		In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task.

	Video: [[media::PCM-RAS.avi]]		Video: [[media:PCM-RAS.avi]]

	==Round-Robin Distributed Q-Learning (2011)==		==Round-Robin Distributed Q-Learning (2011)==

Bortx: /* Partially Constrained Models (2011) */

2011-12-30T11:13:04Z

Partially Constrained Models (2011)

← Revisión anterior		Revisión del 13:13 30 dic 2011
Línea 8:		Línea 8:

	Video: [[media::PCM-RAS.avi]]		Video: [[media::PCM-RAS.avi]]


	==Round-Robin Distributed Q-Learning (2011)==		==Round-Robin Distributed Q-Learning (2011)==

Bortx: /* Q-Learning and the Hose Tranport Application: Videos */

2011-12-30T11:12:15Z

Q-Learning and the Hose Tranport Application: Videos

← Revisión anterior		Revisión del 13:12 30 dic 2011
Línea 7:		Línea 7:
	In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task.		In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task.

			Video: [[media::PCM-RAS.avi]]

Bortx: /* Q-Learning and the Hose Tranport Application: Videos */

2011-12-30T10:43:25Z

Q-Learning and the Hose Tranport Application: Videos

← Revisión anterior		Revisión del 12:43 30 dic 2011
Línea 2:		Línea 2:

	A set of robots is attached to a hose modeled as a line segment between them. Agents are first trained using Q-Learning not to strech the hose above its nominal maximum length and not to collide with each other. The robot carrying the tip of the hose is desired to reach the goal, which is represented as a green dot.		A set of robots is attached to a hose modeled as a line segment between them. Agents are first trained using Q-Learning not to strech the hose above its nominal maximum length and not to collide with each other. The robot carrying the tip of the hose is desired to reach the goal, which is represented as a green dot.

			==Partially Constrained Models (2011)==

			In this video, we show the difference between learning a task in a computationaly expensive environment (GEDS) and first trainning agents on a PCM and then testing learnt policies in the target task.

Bortx: /* Round-Robin Cooperative Multi-Agent Q-Learning (2011) */

2011-12-29T19:53:33Z

Round-Robin Cooperative Multi-Agent Q-Learning (2011)

← Revisión anterior		Revisión del 21:53 29 dic 2011
Línea 4:		Línea 4:


	==Round-Robin ~~Cooperative Multi-Agent~~ Q-Learning (2011)==		==Round-Robin Distributed Q-Learning (2011)==

	In this case, the state is fully observable but the agents do not explicitly coordinate. The reward signal is shared by both agents. Instead of using typical Q-Learning, we use our Round-Robin Cooperative Multi-Agent Q-Learning variation of the algorithm, which forces agents to take actions one by one. After 1000 episodes for each configuration, the system reached an optimal policy fo all configurations.		In this case, the state is fully observable but the agents do not explicitly coordinate. The reward signal is shared by both agents. Instead of using typical Q-Learning, we use our Round-Robin Cooperative Multi-Agent Q-Learning variation of the algorithm, which forces agents to take actions one by one. After 1000 episodes for each configuration, the system reached an optimal policy fo all configurations.