Lower-Limb Exoskeleton With Compliant Actuators: Human Cooperative Control

Active exoskeletons for the lower extremities are increasingly being used in rehabilitation therapy. One of the key areas of research in developing these assistive devices is ensuring safe human-machine interaction, which requires both a safe mechanical system and an effective assistive control framework. Therefore, we present a novel human cooperative control framework for exoskeletons with variable stiffness actuators to assist users during both the swing and stance phases of walking and other motion sequences such as sit-to-stand. The control framework estimates the user's joint torques during the swing and stance phases using an Unscented Kalman Filter (UKF) and inverse kinematics, respectively. Using the Lower-Limb Exoskeleton with Serial Elastic Actuators (L2Exo-SE) as an example, the control approach was validated for its applicability to exoskeletons with compliant actuators. The validation results reveal a reduction of the average user's joint torque during gait by 63.6%-78.4% for the hip and 40.8%-50.2% for the knee compared to non-assisted walking. Furthermore, we introduce an automatic stiffness selection for the serial elasticity of the variable stiffness actuator (VSA) based on the gait phases and the active human joint torque. The stiffness variation increases the physical human-robot interaction during the swing phase while maintaining a high control bandwidth during the stance phase.