Design of a Multi-Joint Passive Exoskeleton for Vertical Jumping Using Optimal Control

Research and the commercial use of exoskeletons that augment human activities are rapidly growing. However, the progress of the two is hindered by the time-consuming and costly process of designing and evaluating the exoskeleton. One of the solutions to reduce both is the use of simulations that model the users, exoskeleton, and their interaction. At the same time, most simulations focus on continuous tasks, such as walking, running, and industrial activities. The augmentation of human capability is essential in fast motion tasks (i.e., jumping, throwing), where the muscles are producing their maximum force. Thus, this study implemented a simulation of passive exoskeleton-human interactions using OpenSim and Moco software for optimal control to find muscle excitation that maximizes vertical jump height. The models include a planar human model with ankle, knee, and hip joints. The muscles were modeled as torque actuators for each joint, with a flexor and an extensor, and passive torques representing each joint's ligaments. The simulation was used to study: a) the effect of different spring stiffness at the knee, hip, and ankle joints and combinations of these joints; b) multi-joints vs. single joints; c) the effect of an elliptic pulley and different initial engagement angle for springs. The results revealed that the jump height increased as the spring became stiffer, up to a maximum point. For a single joint, the knee exoskeleton was the most effective, compared with the hip and ankle joint exoskeletons. The multi-joint exoskeleton was slightly better than the single knee joint. If maximum spring tension is a limiting factor, an elliptic pulley has an advantage relative to a round pulley. An initial angle of engagement (with equal work) other than zero up to approximately 50 degrees does not decrease the jump height.