Control of a Rehabilitation Robotic Device Driven by Antagonistic Soft Actuators

Stroke is becoming a widely concerned social problem, and robot-assisted devices have made considerable contributions in the training treatment of rehabilitation. Due to compliance continuous deformation capacity, rehabilitation driven by soft actuators are attached widespread attention. Considering large output force pneumatic artificial muscle (PAM) biological musculoskeletal structure, an antagonistic PAM-driven robotic device developed. To fulfill need for control proposed device, knowledge-guided data-driven modeling approach used adaptive feedforward–feedback presented ensure motion accuracy under with high frequency. Finally, several simulations experiments carried out evaluate performance developed system, results show that system controller can achieve expected various operations.