A-31 Dynamic Contribution Analysis of Swing Motion with Use of Inertial Sensor Information : Quantification of Racket Speed Generating Mechanism

Abstract

Among high speed swing motions in sports, motion dependent term, which is a dynamics characteristic of multi-body system and is consisting of product sum of angular velocities of individual segments, is the great contributor to generating the tip speeds of hitting tools or distal body segments prior to the times of impact or ball release. Therefore, in order to develop a feedback system for improving the performance, it would be efficient to show speed generating mechanism as well as kinematic and kinetic information to players during immediate feedback. The purpose of this study was to develop a portable analysis system using inertial sensors to evaluate racket head speed generating mechanism during swing motion with consideration of joint torque generating types for immediate feedback in coaching field utilization. Two types of inertial sensors, such as gyro sensor and accelerometer, were used in the proposed system. The gyro sensors were attached on the upper trunk, upper arm, racket, and distal end of forearm segment. And the accelerometers were attached on the known location of upper trunk and racket. The sensor outputs were simulated from experimental data of tennis serve motion collected with a motion capture system (Vicon-MX, 12-camera, 250Hz). The error between estimated and measured accelerations was minimized to identify 12 parameters of initial coordination system of the segments by using function of fmincon (optimization toolbox, MATLAB, Mathworks Inc.) with an object function which minimizes the norm of the acceleration error. By solving the equation relating accelerometer sensor outputs to gyro sensor outputs with consideration of geometric constraint relationships in terms of acceleration, we can obtain initial orientation of individual segments, and can calculate motion data using the initial orientation, gyro sensor outputs, and accelerometer output of the upper trunk. The results of dynamic contribution analysis obtained from the estimated motion data show almost good agreements in terms of head speed generating mechanism.