Effects of control parameters of wearable robotics on muscle activity during assisted elbow flexion

Abstract

One way to provide assistance in a dynamic lifting task is to pre-emptively move the exoskeleton based on a predicted reference trajectory. However, the level of aggressiveness in the prediction (i.e., how far ahead in time) and the exoskeleton's degree of adherence to the reference trajectory (stiffness) are not yet fully understood. This study investigated the effects of stiffness and pre-emptive offset parameters in an impedance-controlled robotic arm on muscle activation and perceived exertion of the user. Thirteen participants were instructed to lift a load equivalent to 15% of their maximal voluntary contracted force in collaboration with a robotic arm with 40°–135° of elbow flexion in 1.12 s. Three levels of stiffness (lower: 0.1 N m deg⁻¹, medium: 0.2 N m deg⁻¹, and higher: 0.31 N m deg⁻¹) and two levels of pre-emptive offsets (shorter: 0.1 s and longer: 0.4 s) were investigated. We found that (1) during 0–0.5 s (acceleration stage) of elbow flexion, a higher stiffness level and a longer pre-emptive offset decreased muscle activity; (2) during 0.5–1 s (deceleration stage) of elbow flexion, medium and higher stiffness with a shorter pre-emptive offset decreased muscle activity; (3) the perceived exertion and assistance of participants were improved with a higher stiffness and a longer pre-emptive offset, whereas cooperation was rated higher at a shorter pre-emptive offset under higher stiffness. This study reveals that the optimal parameters for stiffness and pre-emptive offsets for predictive impedance controls are different for different stages of elbow flexion.