Analysis of Standing Back Somersaults Using Motion Generation and Control Based on Angular Momentum Distribution

Abstract

This work clarifies a possible approach toward the motion generation and control of the standing backward tuck somersault, one of the fundamental maneuvers in gymnastics. We show how to generate various types of backward tuck somersaults by focusing on the angular momentum distribution within the body segments. For this purpose, we use the Relative Angular Acceleration (RAA) controller that was developed for humanoid robot control. The desired trajectories for the center of mass and the system angular momentum are derived from the human performances of the somersault task. Furthermore, the generated motions are analyzed to enhance the performance of the maneuver via computer simulation. Our analysis revealed that the duration of the tuck phase in mid-air and the landing position of the foot play an important role for a stable performance of the task. We achieved similar results as in the field of sports biomechanics.