WITHDRAWN: Stochastic energy saving strategies using machine learning for badminton robots

Using robots as badminton training partners, the robots preferred to be humanoid in shape. However, existing humanoid robots lack the tipping stability required to perform badminton footwork. In this research, to improve the tipping stability of humanoid robots, we proposed to install adsorption mechanisms on the sole. The robot with three degrees of freedom (DOFs) on one leg was developed as a preliminary step for the evaluation of the mechanisms. First, the basic characteristics of the adsorption mechanisms were evaluated. The evaluation showed that the tipping stability of the robot was improved. In addition, three motions were performed with the developed robot: maintaining a standing posture, stepping forward, and stepping backward. These motions were realized by using a combination of feedforward control and proportional-integraldifferential (PID) control. In the forward stepping motion, the robot achieved a quick approach speed of 1.56 m/s. Also, the estimation of the zero moment point (ZMP) during the motion showed that the robot could be prevented from tipping by the adsorption mechanisms.

show abstract

WITHDRAWN: Stochastic energy saving strategies using machine learning for badminton robots

Cited by 2 publications

References 13 publications

A prototype of auto badminton training robot

A prototype of auto badminton training robot

Pneumatic Humanoid Robot with Adsorption Mechanisms on the Sole to Realize a Stepping Motion in Badminton

Contact Info

Product

Resources

About