학술논문

Physical interaction analysis of the human body dorsal region with a robotic manipulator
Document Type
Conference
Source
2021 14th IEEE International Conference on Industry Applications (INDUSCON) Industry Applications (INDUSCON), 2021 14th IEEE International Conference on. :1011-1016 Aug, 2021
Subject
Computing and Processing
Power, Energy and Industry Applications
Robotics and Control Systems
Adaptation models
Solid modeling
Protocols
Biological system modeling
Force
Focusing
Safety
contact
dorsal
MuJoCo
pHRI
physical Human Robot Interaction
safety
Language
Abstract
This article studies the development of a simulated model for the human dorsal. Safety protocols are required in order to avoid injures in humans when in contact with a robotic manipulator. One way of developing these protocols is looking for limit values for the main variables which are responsible for modeling the physics of contact, such as deformation. Physical human-robot interaction researchers have been doing it for each part of the human body. This work used the same process focusing on the dorsal. Developing an anatomical investigation of this human body region made it possible to divide and group parts with similar anatomical compositions. Points were attached to each group creating a 2D map of dorsal composed of 9 of them. Specific stiffness was evaluated mathematically for each point. To put them in the simulation environment, Multi-Joint dynamics with Contact was used. In this software, the dorsal simulated model was put in contact with the LBR iiwa model simulated. It happened under the control of an impedance controller. An iteration process was done using a reference force for contacting interface among the bodies, data of LBR iiwa end-effector displacement, and a linear equation approach relating them to set the stiffness value of each point of the dorsal model according to the one evaluated before. Results showed that the dorsal model and the manipulator stabilized in a common position showing the controller action happened satisfactorily. Geometric restrictions (higher and width) of the completed dorsal model generated some difficulties to stabilize the simulation with the exact stiffness values evaluated. However, the variation of them through the dorsal model was preserved, similarly to the real dorsal stiffness variations.