Sit-To-Stand and Stand-To-Sit Energetics for Assistive Devices and Robotics Design

Abstract

This research presents the development of a Sit-to-Stand and Stand-to-Sit model for regenerative energy recovery with applications in orthoses, protheses and humanoid robot design. Sit-to-Stand and Stand-to-Sit are routine activities and are crucial pre-requisites to walking and running. Determining design parameters for devices which can aid people to perform these activities in an effective manner is a key goal here. MapleSim was used to simulate a 1/10-scale multi-domain model and a nonlinear torque controller was used to control the trajectory profiles of the Sit-to-Stand-to-Sit gait. The model allows accurate simulation of hardware components for use in a future robot. This study addresses the usage in regenerative braking towards sit-to-stand-to-sit and the relationship between Coriolis/centrifugal torque components to inertial and gravitational torque components. This study examines the level of regeneration at ankle, knee and hip. Furthermore, it examines the significance of Coriolis and centrifugal versus inertial and gravitational components of a nonlinear controller in order to determine if these components would be needed in a real robot controller. By applying joint trajectories from human trials it was found that the regenerative effect in the robot model was most significant in the hip and least significant in the ankle. Furthermore, we determined that the Coriolis and centrifugal terms were approximately 1% of the inertial and gravitational terms in the applied nonlinear controller, making them insignificant. We also determined upper bounds for gearing in the joints such that battery autonomy is maximized without encountering motor saturation and inaccurate trajectory following. From these findings, we recommend that robot designs neglect the Coriolis and centrifugal terms and that regenerative hardware be prioritized at the hip.