Resumen
Lower extremity assistive devices (LEADs) have been developed in various fields, such as rehabilitation, military, and industry, in the form of exoskeleton robots or treadmills, and most of them are aimed at supporting muscle strength. However, unlike the aforementioned fields, the objective of LEADs developed in the space field is to provide resistance training to prevent muscle atrophy, which is a problem that arises in astronauts during long-duration space flights. Because the purpose of a LEAD is different from those of systems that are intended for use under Earth gravity (1 g) condition, other factors should be considered for the system design. In this study, the appropriate locations and types of actuators for reproducing the kinematics and muscle-related state variables observed in 1 g normal walking in a microgravity environment were proposed, and the corresponding control inputs obtained using a dynamic optimization simulation method. In detail, two actuation types were proposed, considering the characteristics of a microgravity environment in which both the magnitude of the gravitational acceleration and the ground reaction force were decreased. Moreover, by using the proposed actuating system, the control inputs required to track kinematics data and muscle activity were obtained. A human lower-limb model, with six degrees of freedom, i.e., an 18-muscle model with the pelvis fixed, was used with ideal actuators to apply torques or forces to joints or soles. Dynamic optimization was performed to solve these problems using direct collocation with OpenSim and MATLAB. Using the two proposed types of actuation, the results agreed with the kinematics and muscle activity of 1 g normal walking, and the total joint torques by the muscles also exhibited similar curves to that of the net joint torques under 1 g normal walking. The results of this study suggested an actuation method and its control input that can be used in the design of a LEAD for resistance training in microgravity.