Date

2002

Author

Gürses, Senih

Metadata

Show full item record

Item Usage Stats

132
views

0
downloads

Cite This

The first part of the study deals with developing a protocol to acquire force platform data during quiet stance from healthy subjects. For this purpose, the force platforms in the Biomechanics Laboratory of the Department of Mechanical Engineering, METU are used. A Postural Analysis System is developed in addition to the existing Gait Analysis System which was developed in the same Laboratory, earlier. It is the "first time" that the characteristic frequencies of the human postural sway are identified in this study, through an appropriate signal processing of experimentally recorded data. Moreover, a nonlinear mathematical model is developed and tuned for its system parameters in the frequency domain with respect to the characteristic frequencies identified experimentally. All these lead to the definition of a new physiological transducer for the "first time" in this study, with its working band in the spectrum of the postural sway. The everlasting complex behavior of the signals recorded lead to the development of ideas related to the nonlinear system dynamics. The nonlinear dynamical tools used in this study, such as phase portraits directly constructed from the experimentally recorded signals, power spectra of the signals, Poincare sections inobtained from constructed phase planes of the signals, show invaluable hints about the deterministic characteristics of the oscillations. These characteristics still remain to be a great debate between the groups discussing the stochastic characteristics of the postural sway either on behalf or against. Nevertheless, a more direct and certain evidence about the chaotic oscillations of the postural sway arise at the end of the signal processing for calculating the largest Lyapunov exponent from the experimentally recorded time series. For this purpose, a number of software and a modified Wolfs method of calculating largest Lyapunov exponent from a time series at the Poincare section was developed and used for the "first time" in this study. The nonlinear mathematical model developed points out and enriches the reasons of the results that are directly extracted from the experimental recordings through simulations. The simulations make it possible to define the physiological ranges for the system parameters, which are also thought to be a very important aspect of the present research, as it gives opportunity to differentiate physiological from pathological. Although, only a physiological system identification was aimed originally, the same Postural Analysis System developed at the Biomechanics Laboratory of the Department of Mechanical Engineering, METU is used to create a huge series of database for the central and peripheral nervous system disorders. It is the future work that the developed models and nonlinear dynamical tools would be used to identify pathological cases and assist for the therapy by introducing new perspectives as well as can be used in a more unified understanding of the governing rules of the human brain.

Subject Keywords

Posture, Human beings, Human mechanics, Motor ability, Postural sway, Mathematical modeling, Characteristic frequencies, Sensorial dynamics, Sensor thresholds, Control strategies, Postural dynamics, Postural stability, Nonlinear dynamics, Phase portraits, Physiological system parameters, Poincare sectioning, Lyapunov exponents, Time series, Chaotic dynamics, Simple pendulum revisited, Determinism versus probability, Evolutionary biology, Motor control, Shear deformation sensor

URI

https://hdl.handle.net/11511/12870

Collections

Graduate School of Natural and Applied Sciences, Thesis