Posture control and balance for quality of walking in rehabilitation
Date of Issue2016
School of Mechanical and Aerospace Engineering
Robotics Research Centre
Impairments such as spinal cord injury (SCI), stroke or aging can cause deficits in motor control and muscle weakness. The deficits will affect walking ability and the individuals’ ability to live independently. Rehabilitation process can help the impaired individuals improve their walking ability by large volume of repetitive walking. Nevertheless, many subjects still require lower-limb assistive devices such as lower-limb orthosis and lower-limb robotic exoskeleton to compensate for unrecoverable functions. However, most current lower-limb assistive devices do not provide adequate assistance that is needed to maintain postural balance. Additional stabilizing forces are required to assist postural balance. The stabilizing forces can be generated by using upper-limb to control a passive balance stabilizing device (without electric actuator) such as a crutch or walker. However, poor coordination and control of the lower-limb-assistive devices and balance stabilizing device may lead to high energy cost due to high force load on upper limb and it may even destabilize instead of stabilizing the user. It is therefore necessary to find ways to improve user-independence which requires human effort to support balance during walking with lower-limb assistive devices. Thus, the research work reported in this thesis focuses on developing methods of coordination and control of human posture and the stabilizing devices to make sure that they are working closely and well together to reduce human effort. The research work covers three aspects: human posture and passive stabilizing device control, trajectory generation and human-intention recognition, and development of active balance stabilizer mechanism. A torque-driven, five segments (two thigh segments, two shank segments and one trunk segments) forward dynamics simulation model is first used to study the human posture and passive stabilizing device control. Based on the stability analysis, crutch support forces, support positions and CoM states (including position, velocity and acceleration) are identified as three main effectors. The forward dynamics simulation results of this study provide the suggestion on the correct ways of posture control and manipulating the balance stabilizers to achieve a stable walking. Based on the simulation study, a human-intention recognition algorithm is developed to coordinate the control of the stabilizing device and lower-limb exoskeleton. The algorithm is developed based on a stability index named Extrapolated Center of Mass (XcoM). The motion of the lower-limb robotic exoskeleton will only be triggered when a safe threshold of the XcoM is reached. A human-like trajectory generation method is developed to control the movements of the robotic exoskeleton. A robotic exoskeleton type lower-limb assistive device with lateral body weight shifting ability is designed and built. The developed exoskeleton is proven to be able to trigger a stable swing motion in terms of XcoM stability criteria and reduce the human effort by 50% compared with the exoskeleton without lateral balance shifting ability. Lastly, an additional active balance stabilizer is used to further improve user-independence and balance when walking with lower-limb assistive device. It is a stick type device which has a point contact with the ground and is controlled by two electric motors. A minimum jerk trajectory generation method based on polynomial function is used to synchronize the motion of the lower-limb exoskeleton and balance stabilizer mechanism. A balance filtering algorithm using whole-body Center of Gravity (CoG) Jacobian is developed to filter the unstable part of the trajectory based on the Zero Moment Point (ZMP) principle. Tests on healthy subjects have been successfully conducted to prove the concept of the proposed research work. This work demonstrated the possibility of significantly reducing the arm load by 83% with the support from the balance stabilizer comparing with walking with only robotic exoskeleton. In all, the stabilizer mechanism is able to maintain stability, facilitate body weight shifting, reduce human effort and the number of helpers required during walking. In summary, methodologies on reducing the human effort when walking with robotic exoskeleton were proposed and their applicability was validated by experiments with healthy subjects.