Optimal Selection of Motors and Transmissions in Back-Support Exoskeleton Applications

Because of the central role the actuator unit (electrical motors and transmission parts) has in wearable robots, improving the performance of the torque/force control system is vital, particularly for exoskeletons. This paper proposes an optimal approach to the selection of the main components of an actuation system (brushless DC motor and gearbox transmission) to be used in a back-support exoskeleton, but the principles can be extended and applied to other types of exoskeletons. To perform the optimization, an analytical model based on the dynamics of human-robot interaction has been developed. Moreover, to incorporate the weight of the actuator in the optimization framework, a mathematical relation between the weight and technical characteristics of the components, based on the polynomial regression technique using the low-discrepancy sequences method are developed. Consequently, the optimization criteria in terms of the closed-loop system frequency bandwidth, system power consumption and the weight of the components are formulated by imposing technical constraints on simulation parameters. The optimization results demonstrate two possible actuator combinations. Subsequently, the selected actuator components are evaluated in a lifting scenario by means of a linear quadratic regulator (LQR) controller with double integral action. Extensive simulation results in terms of the control frequency bandwidth, torque tracking control, current produced by motors and system robustness with respect to external disturbances are presented and discussed to make comparisons between the possible combinations of the components and their feasibility in the back-support exoskeleton applications.