Studies on estimation of three-dimensional rotational axis in figure skating jump by Iterated Extended Kalman filter and quaternion using inertial magnetic field sensors

Abstract

This paper proposes the estimation method of three-dimensional rotational axis in figure skating jump by iterated Extended Kalman filter and quaternion using inertial and magnetic field sensors. This method estimates the posture information to global coordinate system from the sensor coordinate system and the posture information to rotational axis coordinate system from the sensor coordinate system using the iterated Extended Kalman filter, and the direction component represented by quaternion was estimated by using inertial and magnetic field sensors. We conducted the measurement experiment the triple flip jumps and the quadruple salchow jumps by the high-level figure skater using inertial and magnetic field sensors, and we estimated the three-dimensional rotational axis in the triple flip and quadruple salchow jumps. These results indicated the change of rotational axis during jump by the normalized vector and the projection drawing calculated by the estimated three-dimensional rotational axis, and we indicated the high reproducibility in the triple flip jumps and the difference of characteristics by the difference in success of jump in the quadruple salchow jumps.