The Quasi-Stiffness Control Parameter and Soleus Muscle Activity Relationship During Powered Ankle Exoskeleton Walking

Abstract

Active exoskeletons are driven by controllers, but the precision required to assist humans efficiently is still unclear. Exoskeleton controllers store commands to aid people during walking, and changes in the exoskeleton control parameters, such as quasi-stiffness, can influence gait characteristics and muscle activation. Exoskeleton ankle quasi-stiffness is how much the exoskeleton ankle will rotate when the user moves their ankle and can be defined as the relationship between exoskeleton ankle torque and angle. This study evaluated the relationship between changes in quasi-stiffness coefficient and muscle activation. Twelve participants (6 female, 6 male, average age 23.08 ± 3.50) walked on a treadmill at 1.25 m/s while wearing a bilateral powered exoskeleton (Dephy ExoBoot) at seven quasi-stiffness values. Soleus surface electromyography (EMG) was collected. Since the exoskeleton assists during the stance phase of the gait cycle, data was analyzed during dorsiflexion (35-60% of the gait cycle). Mean peak values were calculated for torque and EMG as the peak value during dorsiflexion for each gait cycle, averaged across each quasi-stiffness value. A linear mixed-model regression model found an effect of mean peak torque on mean peak EMG. It was found as mean peak torque increases due to increased quasi-stiffness, mean peak soleus activity decreases. This result indicates that as an exoskeleton provides more assistance, through increasing the quasi-stiffness coefficient, users benefit from the exoskeleton assistance and reduce their soleus muscle activation. This information can aid the development of new co-adaptive controllers to ensure exoskeletons can enhance or assist user’s capabilities effectively.