Date of Award

2008

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Bioinformatics & Computational Biology

First Advisor

Dr. Kamran Iqbal

Abstract

Biomechanical movements with multiple degrees of freedom exhibit complex coordination and control mechanisms in the presence of physiological constraints. A human body biomechanical model which can mimic diverse 3D physiological movements is a challenging task. In this study, we develop a mathematical framework to study human sit-to-stand (STS) movement using 2D biomechanical models. This analytical framework models the physiological system for feedback controller design. A 2D 4-segment biomechanical model with optimal feedback controller design provides a motivation to extend this scheme for analytical 3D bipedal modeling. We used a newly developed mathematical modeling toolbox (Maple’s DynaFlexPro) to generate 3D 8-segment bipedal models that includes 2 feet, 2 lower leg segments (shanks), 2 upper leg segments, (thighs) a pelvic and a HAT segment. This modeling scheme has 7 joint angles in its sagittal plane and 3 joint angles in its frontal plane with their respective torque drivers. We use experimental data of STS maneuver to synthesize a linear reference model for joint angles. This modeling scheme has several possibilities of connecting a foot joint with the ground and accordingly imposes a varying number of holonomic constraints to the model. A nonlinear model with 0 and 6-degrees of freedom joints in the right and the left foot respectively, linearizes to a 26 th order model with 6 position and velocity based holonomic constraints. We extend this model to 32 nd order model with the addition of pelvic joint variables to analyze the symmetrical and asymmetrical behavior of both right and left lower extremities. We develop linear controllers through H 2 and H ∞ designs for decoupled constrained and unconstrained systems respectively for the control of nonlinear model. We analyze the STS movement with physiologically relevant controller designs based upon kinematic and kinetic variables, including decoupling of rotational and translational variables, decoupling of sagittal and frontal planes, forward thrust phase, and initiation torques at the joints. We provide experimental results for kinetic variables underlying STS maneuver from 6 males and 5 females to compare simulation results for validating our model. The proposed analytical framework can be further extended to complete musculoskeletal dynamics, and shows its applicability to the study task specific applications in kinesiology, ergonomics, and rehabilitation robotics.

Download

Share

COinS