The Role Played by Mass, Friction, and Inertia on the Driving Torques of Lower-Limb Gait Training Exoskeletons

Abstract

Lower-limb gait training exoskeletons are extraordinary tools used to reduce the burden of locomotor impairments in patients with neurological diseases. However, the transparent operation and backdrivability of such systems still needs to be improved. Moreover, it is not completely understood how the mechanical design of the robot can interfere with the user's gait pattern. Aiming to address these shortcomings, we investigate the required driving torques and mechanical power to move the legs under a wide range of actuator's mass, inertia and friction and thigh/shank lengths. We used the ExoRoboWalker, a six-degree-of-freedom lower-limb exoskeleton, to build a framework model based on the double-pendulum approach integrated with the actuators' mechanical impedance. Decoupled joint apparent inertia and the Rayleigh's dissipation function were introduced to the robot's Lagrangian to consider the effects of gearhead ratio and joint friction in the model. Firstly, it is presented the isolated effect of such variables on the required driving torques of the system. The oscillation frequency for the minimum joint torque was severely affected by variations of inertia, friction, and links length. Secondly, the combined effect of the actuator's mass, inertia and friction reveled that a heavier exoskeleton with low-ratio transmission required less torque and mechanical power than a lighter one with greater reduction ratio depending on the oscillation frequency, which is remarkable. These findings have important implications for new designs of lower-limb gait training systems.