Locomotion Learning in Modular Self-reconfigurable Robots

Advisor Information

Prithviraj Dasgupta

Location

UNO Criss Library, Room 231

Presentation Type

Oral Presentation

Start Date

3-3-2017 11:30 AM

End Date

3-3-2017 11:45 AM

Abstract

We study the problem of adaptive locomotion learning for modular self-reconfigurable robots (MSRs). MSRs are mostly used in unknown and difficult-to-navigate environments where it can take a completely new shape to accomplish the current task in hand. Therefore it is almost impossible to develop the control sequences for all possible configurations with varying shape and size. The modules have to learn and adapt their locomotion in dynamic time to be more robust in nature. In this project, we propose a Q-learning based locomotion adaptation strategy which balances the exploration versus exploitation in a more sophisticated fashion. We have applied our proposed strategy on ModRED modular robot within Webots simulator environment.

Additional Information (Optional)

Winner of Outstanding Graduate Oral Presentation

COinS
 
Mar 3rd, 11:30 AM Mar 3rd, 11:45 AM

Locomotion Learning in Modular Self-reconfigurable Robots

UNO Criss Library, Room 231

We study the problem of adaptive locomotion learning for modular self-reconfigurable robots (MSRs). MSRs are mostly used in unknown and difficult-to-navigate environments where it can take a completely new shape to accomplish the current task in hand. Therefore it is almost impossible to develop the control sequences for all possible configurations with varying shape and size. The modules have to learn and adapt their locomotion in dynamic time to be more robust in nature. In this project, we propose a Q-learning based locomotion adaptation strategy which balances the exploration versus exploitation in a more sophisticated fashion. We have applied our proposed strategy on ModRED modular robot within Webots simulator environment.