Cable-Driven Flexible Exoskeleton Robot for Abnormal Gait Rehabilitation

The number of people with abnormal gait in China has been increasing for years. Compared with traditional methods, lower limb rehabilitation robots which address problems such as longstanding human guidance may cause fatigue, and the training is lacking scientific and intuitive monitoring data. However, typical rigid rehabilitation robots are always meeting drawbacks like the enormous weight, the limitation of joint movement, and low comfort. The purpose of this research is to design a cable-driven flexible exoskeleton robot to assist in rehabilitation training of patients who have abnormal gait due to low-level hemiplegia or senility. The system consists of a PC terminal, a Raspberry Pi, and the actuator structure. Monitoring and training are realized through remote operation and interactive interface simultaneously. We designed an integrated and miniaturized driving control box. Inside the box, two driving cables on customized pulley-blocks with different radii can retract/release by one motor after transmitting the target position to the Raspberry Pi from the PC. The force could be transferred to the flexible suit to aid hip flexion and ankle plantar flexion. Furthermore, the passive elastic structure was intended to assist ankle dorsiflexion. We also adopted the predictable admittance controller, which uses the Prophet algorithm to predict the changes in the next five gait cycles from the current ankle angular velocity and obtain the ideal force curve through a functional relationship. The admittance controller can realize the desired force following. Finally, we finished the performance test and the human-subject experiment. Experimental data indicate that the exoskeleton can meet the basic demand of multi-joint assistance and improve abnormal postures. Meanwhile, it can increase the range of joint rotation and eliminate asymmetrical during walking. © 2021, Shanghai Jiao Tong University and Springer-Verlag GmbH Germany, part of Springer Nature.