A humanoid robot with the intelligence of playing a swing

A humanoid robotic system mimics human behaviors with human-like intelligence. In order to give robots human-like dexterously operational skills, such as walking, jumping, running, etc., researchers have been working on implementing human-like intelligence into machines. This paper presents a humanoid robot with the intelligence of playing a swing so that the robot is able to simulate playing strategies of human beings on a swing system. Based on observation and analysis of human postures while playing a swing, a simplified humanoid robot with a two-link architecture was designed and developed. Those two links represent the main part of the swing system and the legs of the human, respectively. Because the number of actuator is less than the total number of the degrees of freedom, the whole swing system becomes an under-actuated system. Fuzzy control approach was chosen to implement swing strategies with human intelligence. Both computer simulations and experimental results successfully demonstrate human-like swing control techniques.