A Novel Pneumatic Artificial Muscle -Driven Robot for Multi-joint Progressive Rehabilitation

Xingxing Guo,Quan Liu,Jie Zuo,Wei Meng,Qingsong Ai,Zude Zhou,Wenjun Xu
DOI: https://doi.org/10.1109/devlrn.2018.8761014
2018-01-01
Abstract:Due to the bio-mechanical characteristics and inherent compliance, pneumatic artificial muscles have been widely applied in rehabilitation robotic field. However, the most existing multi-joint rehabilitation robots have the disadvantages of bulky facilities, low utilization rate and high cost; while some rehabilitation robots with simple mechanism are only suitable for a specific joint rehabilitation. This paper presents a single degree of freedom rehabilitation robot with progressive adjustation ability, which can provide suitable assistance for different patient's injury site. By introducing the joint motion radius element, the robot's mechanical parameters, fixed position, drive unit's overhanging state can be adjusted to provide the required range of motion and assistance torque to adapt to each recovery period during the whole rehabilitation process. After the kinematics and dynamics model of the joint mechanism is established, a modified sliding mode control method based on RBF neural network is utilized to compensate the system disturbance and guarantee the robust stability of the control. The experimental results show that the adopted algorithm achieved better control performance than the traditional sliding mode control method, which is suitable for the rehabilitation training of patients during the entire progressive rehabilitation periods.
What problem does this paper attempt to address?