Study of Human Motion Control with a Physiologically Based Robotic Arm and Spinal Level Neural Controller

Chou, C.P. (1996) Study of Human Motion Control with a Physiologically Based Robotic Arm and Spinal Level Neural Controller. Doctoral thesis, University of Washington.

Full text not available from this repository.


The goals of this research are: 1) to apply knowledge of human neuro-musculo-skeletal motion control to build a physiology based neural controller and biomechanics based "anthroform" robotic arm system, and 2) to utilize the neural controller and anthroform arm system to study some controversial issues and to predict new phenomena of the human motion control system. A physiologically analogous artificial neural network controller, a feed-forward controller, and two anatomically accurate robotic arms (the Anthroform Arm and the testing elbow) are implemented and applied in this study. In order to build the physical arm systems to have mechanical properties as close as possible to the human arm, McKibben pneumatic artificial muscles, force sensors, and mechanical muscle spindles are integrated in the systems with anatomically accurate muscle attachment points. <p> In addition, the neural controller emulates the behavior of spinal segmental reflex circuitry in real time, which includes motoneurons, interneurons, and Ia and Ib afferent feedbacks. The feed-forward controller emulates the inverse kinematics and inverse dynamics functions of higher central nervous system to convert conceptual movement parameters to spinal neural activation patterns. <p> Systematic experiments of elbow posture maintenance and voluntary fast flexion movement are performed and compared to physiological experimental data. New experiments are performed in which responses to torque perturbation are measured when selected afferent pathways are blocked. A "covariance diagram" is introduced. And a linear model is used to help to analyze the roles of system components and the stability. The results show: 1) the neural controller and testing elbow system is capable of similar performance in posture maintenance as the human elbow. 2) Muscle co-contraction and Ia afference with gamma motoneuron excitation are two effective ways to increase joint stiffness and damping, which in turn reduces the mechanical sensitivity of the joint to external perturbation and shorten the settling time of the system. 3) Feed-forward inverse dynamics information is important to performing voluntary fast movement. And 4) feedback delay limits the maximum loop gain of the closed-loop controlled system to maintain stability and still produce enough additional torque to resist external perturbations during voluntary movement. A higher level "nearly open loop" predictive feedback mechanism is proposed for solving this high-loop-gain instability problem. Physiological experiments are encouraged to verify the above phenomena in the future.

Item Type: Thesis (Doctoral)
Subjects: A Neural Engineering > A Neural Engineering (General)
Divisions: Department of Electrical Engineering
Depositing User: Jeffrey Herron
Date Deposited: 07 Jul 2015 21:24
Last Modified: 07 Jul 2015 21:24

Actions (login required)

View Item View Item