This dissertation extends the characterization of human arm impedance control beyond the measurement of unloaded postural stiffness. Measurement of the dynamic components of postural impedance (mass and damping), characterization of postural stiffness modulation under various cognitive conditions and in the face of initial static loading, and the development and initial testing of an identification methodology for the measurement of impedance in the voluntary case, in which the human arm is in motion.; The design and construction of a two-degree-of-freedom, direct-drive robotic manipulator with a force-sensing handle as a perturbation and measurement device is described. We develop an identification methodology which includes calculation of dynamic, as well as static, components of impedance. The relationship between human arm joint stiffness and damping is examined, and the most likely hypothesis relates joint damping to both joint stiffness and joint inertia, rather than to joint stiffness alone. Two-joint impedances, i.e. impedances associated with muscles connected across both the elbow and shoulder joints, are found to play a relatively smaller role in damping than in stiffness.; The ability of the CNS to modulate the stiffness component of impedance cognitively and under initial static bias forces is examined. In response to cognitive cues, stiffness shapes change most dramatically, in comparison to shape and orientation. In the case of initial bias forces, shifts in the human arm endpoint's spring center, or equilibrium point, corresponding to the bias force directions were observed, along with shape changes on the order of those observed in the case of cognitive modulation. The observation of regular shifts in the spring center, or equilibrium position, of the human arm to confirms the validity of the impedance control hypothesis in the loaded postural case.; A methodology is proposed for the measurement of human arm impedance while the arm is in motion, i.e., voluntary impedance. A force-pulse-based method of human arm impedance identification gives results comparable to those yielded by a position-control approach in the postural unloaded case. (Abstract shortened with permission of author.)
展开▼