Date of Award

Fall 2009

Document Type


Degree Name

Doctor of Philosophy in Mechanical Engineering - (Ph.D.)


Mechanical and Industrial Engineering

First Advisor

Zhiming Ji

Second Advisor

R. S. Sodhi

Third Advisor

Bernard Koplik

Fourth Advisor

Ian Sanford Fischer

Fifth Advisor

Richard A. Foulds


While most take their ability to walk for granted, some are unable to walk due to any number of pathologies, such as traumatic brain injury (TBI), spinal cord injury (SCI), and Parkinson's disease. Decreased activity has been shown to be associated with rapidly deconditioning. Rehabilitation techniques that afford patients the ability to begin reconditioning through walking sooner may ultimately enhance their return to a better quality of life. To assist the functional recovery of such patients, an appropriate afferent input to the spinal cord will help in the therapy of the patient. Manual training is labor intensive, costly, and ergonomically unfavorable and tiring to the trainer, which will make training sessions short. Task-specific exercises delivered by robotic devices have registered success in reducing impairment and increasing motor power. Advantages of using robotic devices over the manual training include: reproducibility of the movement in a physiological manner, prolongation of the training sessions. However, several shortcomings exist in the existing devices. The reliability of the hardware and software is critical for their safe operation. These devices are expensive and only available in some large rehabilitation research institutions, which limit many disabled people to get the therapy they require.

A gait pattern generation system was developed, which uses close-chain linkage mechanisms to guide the legs with coordinated movement of the leg joints to follow normal physiological gait pattern. The natural kinematics constraints of the mechanisms produce trajectories that limit the joints' range of motion (ROM) thus improving the gait training safety. The movement of the input links of gait generation mechanisms must be controlled to provide proper timing of both stance and swing phases during a training session. This research describes the development of two methods for the motion control and coordination of the linkage mechanisms. First, the desired motion profile of the input crank of the linkage mechanism is obtained in the form of continuous function of time. The first method controls the input crank of the linkage mechanism directly with a servomotor based on the derived continuous motion profile. The coordination of the mechanisms for the two legs is carried out in the motion control software. The second method proposes to achieve both timing and coordination through mechanical means so that the system works under a simple speed control. A combination of cam and planetary gear mechanisms is synthesized for this approach. The procedures for obtaining the gears ratio of the planetary mechanism, the follower motion profile and the cam profile of the cam mechanism are discussed. Computer simulation is used to demonstrate the correct timing of the ankle movements on the closed path with the two methods. It also shows that the corresponding hip and knee joints follow normal physiological gait pattern.