Analysis of Body Kinematics and Force Modulation

Sammanfattning: To analyse the connection between kinematics and stabilizing forces related to human postural dynamics, fourteen normal subjects were investigated with posturography. Stimulation signals according to Pseudo-Random- Binary-Sequence patterns were applied to each subject. The stimulation consisted both of calf vibration and of combined calf vibration and galvanic stimulation to the inner ear. The tests were made with eyes open and with eyes closed. The stabilizing forces and torques against the floor were recorded by a force platform, while the body motions were recorded by an ultrasound system Zebris for measuring 3D-motion. System identification routines were employed to derive connections between the stimuli, the movements, the forces and the torques. Two subspace model identification algorithms were evaluated on the measurement data and the stimulated movement patterns were investigated using variance and correlation analysis. The results from the evaluation showed that the Multivariable Output-Error State Space Algorithm (MOESP) of Verhaegen et al is superior to the Numerical algorithms for Subspace State Space System IDentification (N4SID) in modeling performance. The investigation of movement patterns proved that standing humans exposed to calf vibration and vestibular galvanic stimulation under the test conditions in this study sway as single-segment pendulums.