The study of posture control system has important implications for sports, rehabilitation medicine, and ergonomics. To gain an understanding of internal mechanism of the posture control system, researchers have considered a human posture control system as a feedback system of which a controlled object is an upright body. This thesis is concerned with the two-dimensional automatic control modeling of a posture control system.
The first part of this thesis is devoted to developing a computer-aided instrumentation system for the automatic measurement of postural sway. We used only the Kistler force platform and amplifiers, and developed the data acquisition and analysis software in $TURBO-C^？$ environment using an IBM PC. This system was used to obtain the continuous records of center of pressure location at a rate of 100 per second. The major advantage of the system consists in the automatic measurement of human movement and the high sampling rate enough to study human postural stability.
The second part is concerned with parameter estimation using least squares' method and stability criterion of the identified system. A program in $TURBO-C^？$ environment for the parameter estimation is described. A $MATLAB^？$ with $SIMULINK^{TM}$ program for root-locus plot is also explained.
The last part is devoted to developing posture control model based on two-dimensional feedback control theory. Human postural characteristic was investigated in five healthy subjects. Tests were performed with eyes open and eyes closed. After 5 seconds' quiet standing, each subject was pulled forward by 30mm at his pelvis height unexpectedly and then released. Postural sway was measured over 20 seconds at a rate of 100 per second. Transfer functions to represent posture control characteristic were identified by least squares' method. These showed good results of the model fitness, predictability, and stability. The response of eyes closed condition to perturbation is more oscillatory than that of eyes open condition.
