POSTURE CONTROL OF A FREE FALLING ROBOTIC CAT FOR SOFT LANDING USING REINFORCEMENT LEARNING
Robot’s posture control ability in the air is required when designing advanced robots that can run, jump and land, which can perform tasks in workplaces where ordinary robots cannot go. Using such a robot could afford human safety as well as cost reduction. In this paper, we describe the control method of robot’s posture in its falling for the safe landing using reinforcement learning (RL). The posture control ability for safe landing is inspired by cats. Although a cat is dropped from the upside down posture, she can flip her body so that she can land on her feet. The robotic cat that we simplified from the real cats consists of two rigid columns and has two actuators at the joint to assert torque inputs. The controller is obtained via Sarsa. Based on simulation results, we concluded that the posture control ability of a two column robot can be realized using RL.
reinforcement learning biologically inspired robots
J.J.Lee B.G.Shin J.J.Kim
KAIST,South Korea
国际会议
2008年拟人系统国际会议(2008 International Conference on Humanized Systems )(ICHS’08)
北京
英文
2008-10-18（万方平台首次上网日期，不代表论文的发表时间）