Exoskeleton robots are frequently applied to augment or assist the user's natural motion. Generally, each assisted joint corresponds to at least one specific motor to ensure the independence of movement between joints. This means that as there are more joints to be assisted, more motors are required, resulting in increasing robot weight, decreasing motor utilization, and weakening exoskeleton robot assistance efficiency. To solve this problem, the design and control of a lightweight soft exoskeleton that assists hip-plantar flexion of both legs in different phases during a gait cycle with only one motor is presented in this paper. Inspired by time-division multiplexing and the symmetry of walking motion, an actuation scheme that uses different time-periods of the same motor to transfer different forces to different joints is formulated. An automatic winding device is designed to dynamically change the loading path of the assistive force at different phases of the gait cycle. The system is designed to assist hip flexion and plantar flexion of both legs with only one motor, since there is no overlap between the hip flexion movement and the toe-offs movement of the separate legs during walking. The weight of the whole system is only 2.24 kg. PD iterative control is accomplished by an algorithm that utilizes IMUs attached on the thigh recognizing the maximum hip extension angle to characterize toe-offs indirectly, and two load cells to monitor the cable tension. In the study of six subjects, muscle fatigue of the rectus femoris, vastus lateralis, gastrocnemius and soleus decreased by an average of 14.69%, 6.66%, 17.71%, and 8.15%, respectively, compared to scenarios without an exoskeleton.
Keywords: gait event detection; iterative control; lower limb assistance; soft exoskeleton; time-division multiplexing.