a fuzzy gait phase detection for rehabilitation of hemiplegic patients with a hip exoskeleton robot

Movement disorders, caused by a variety of factors like aging, stroke, and spinal cord injuries, may largely decrease the quality of life of the affected people. compared to traditional methods, robotic neurological rehabilitation methods have shown promising results in improving the performance of the patients. in this paper, a novel algorithm is proposed for rehabilitation of hemiplegic stroke patients, using a hip exoskeleton robot. the proposed algorithm includes a fuzzy inference engine that receives the angle and the angular velocity of the hip joint and estimates the current gait phase of the user. upon detecting the initiation of the swing phase of the affected leg, the corresponding hip motor of the robot applies a predefined assistive torque for a given period of time. when the assist period is finished, a zero-impedance controller will take over the control of the robot with a minimum contact force. the performance of the proposed algorithm is experimentally evaluated using a prototype of a custom-made hip exoskeleton. the experimental results verify the successful performance of the designed fuzzy inference engine in proper timing of the assistive torque, by the precise detection of the correct gait phase of the hemiplegic stroke patients.