Gait locomotion generation and leg muscle evaluation for overground walking rehabilitation robots
Date of Issue2012
School of Mechanical and Aerospace Engineering
Robotics Research Centre
The number of people with difficulty in walking has increased with ageing issues worldwide. It is possible to regain walking ability after persistent locomotion training. In decades, robotic devices have been suggested to enhance motor recovery by replicating clinical manual-assisted training, which is labour-intensive and hard to persistent. This thesis presents a research work on a robotic-assisted training system based on overground walking (OGW), instead of the treadmill training. Proposed motion generation, control and evaluation of the lower limb rehabilitation are specified to the OGW training system for a more natural and effective gait training. In the present work, human motor control is first investigated to avoid a totally passive walking exercise on robotic system. In order to enhance effect of the active training, incorrect muscle activations on lower limbs for the pathological gait are identified. After investigating human motor control for one joint, motion generation algorithms of gait training are proposed for coordinating all joints on the pelvis and lower limbs. For different individual gaits and physical dimensions, an adjustable motion generation algorithm is proposed. Then, the intelligent control methods are also proposed for better tracking the generated motion under different walking speeds. All these proposed motion generation and control methods are implemented on an OGW gait system: NaTUre-gaits (Natural and TUnable rehabilitation gait system). The synchronized control is implemented for the modular systems, i.e. pelvic arm, robotic orthosis and mobile platform. Comparisons of lower limb muscular activation walking with and without the robotic device are studied and discussed. Results of successful clinical-trials show that the performance of the device is able to provide comparable performance obtained by manual assistance by gait rehabilitation training. Further quantifying assessment of muscle dysfunction of human gait is investigated, which works as an effective tool and contributes to qualified evaluation and bio-feedback for gait rehabilitation. In conclusion, the present work has dealt successfully with a full-range of new robotic-assisted gait rehabilitation methodology, which includes human gait locomotion study, development of motion generation and control strategy, system implementations, clinical applications, and quantitative assessment for an OGW rehabilitation training robot.