The Study on Optimal Gait for Five-Legged Robot with Reinforcement Learning

The research of legged robot was rapidly developed. It can be seen from recent ideas about new systems of robot movement that take ideas from nature, called biology inspired. This type of robot begins replacing wheeled robot with various functions and interesting maneuvers ability. However, designers should decide how many legs are required to realize the ideas. One of the ideas that are rarely developed is odd number of legs. This research focused on five legs robot that inspired from starfish. To realize the intelligent system in robot that does not depend on the model, this research used reinforcement learning algorithm to find the optimal gait when robot is walking. In order to achieve this goal, trial and error have been used to provide learning through an interaction between robot and environment based on a policy of reward and punishment. The algorithm is successfully implemented to get the optimal gait on a five-legged robot.

[1]  Erden,et al.  FREE GAIT GENERATION WITH REINFORCEMENT LEARNINGNEXT TERM FOR A SIX-LEGGED ROBOT , 2008 .

[2]  Youcef Zennir,et al.  Multiactor approach and hexapod robot learning , 2005, 2005 International Symposium on Computational Intelligence in Robotics and Automation.

[3]  P. Couturier,et al.  Distributed Reinforcement Learning of a Six-Legged Robot to Walk , 2003, 2003 4th International Conference on Control and Automation Proceedings.

[4]  Alfred A. Rizzi,et al.  Gait Regulation and Feedback on a Robotic Climbing Hexapod , 2006, Robotics: Science and Systems.

[5]  George A. Bekey,et al.  Gait Adaptation in a Quadruped Robot , 2002, Auton. Robots.

[6]  Peter Stone,et al.  Policy gradient reinforcement learning for fast quadrupedal locomotion , 2004, IEEE International Conference on Robotics and Automation, 2004. Proceedings. ICRA '04. 2004.

[7]  K. Inagaki.  Gait study for hexapod walking with disabled leg , 1997, Proceedings of the 1997 IEEE/RSJ International Conference on Intelligent Robot and Systems. Innovative Robotics for Real-World Applications. IROS '97.

[8]  Shigenobu Kobayashi,et al.  Reinforcement learning of walking behavior for a four-legged robot , 2001, Proceedings of the 40th IEEE Conference on Decision and Control (Cat. No.01CH37228).

[9]  Katsuhiko Inagaki,et al.  Adaptive wave gait for hexapod synchronized walking , 1994, Proceedings of the 1994 IEEE International Conference on Robotics and Automation.

[10]  Alberto Rovetta,et al.  Structure Design and Locomotion Analysis of a Novel Robot for Lunar Exploration , 2007 .

[11]  Tim D. Barfoot,et al.  Experiments in learning distributed control for a hexapod robot , 2006, Robotics Auton. Syst..

[12]  Manuela M. Veloso,et al.  An evolutionary approach to gait learning for four-legged robots , 2004, 2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (IEEE Cat. No.04CH37566).