Journal article 198 views 94 downloads
Dynamic manipulation of pneumatically controlled soft finger for home automation
Measurement, Volume: 170, Start page: 108680
Swansea University Author: Shuai Li
PDF | Accepted Manuscript
© 2020. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/Download (7.08MB)
Soft robots have the advantage of inherent flexibility, adaptability, compliance, and safety in human interaction, and therefore attracted significant research attention in recent years. They have found interesting applications in industrial automation where soft robotic hands are fitted as end-effe...
Check full text
No Tags, Be the first to tag this record!
Soft robots have the advantage of inherent flexibility, adaptability, compliance, and safety in human interaction, and therefore attracted significant research attention in recent years. They have found interesting applications in industrial automation where soft robotic hands are fitted as end-effector on traditional rigid robotic arms to handle delicate objects. Their inherent compliance with the shape of the object reduces the complexity of sensing and actuation mechanisms required for the safe operation of traditional robotic hands. They also have the potential application in the home automation, since the operation of robots in indoor environment impose a stringent requirement on safety and compliant design. Despite this, the dynamic manipulation of soft robots remains challenging because their inherent flexibility makes their mathematical model highly nonlinear. Existing works either use model-free control, e.g., PID, which owing to its general formulation, does not account for the peculiarity of soft robots, or they use the Finite-Element-Method based approach, which, apart from being computationally expensive, requires an exact model of the soft robots. In this paper, we take a holistic approach by first developing a low-order approximate mathematical model for computational efficiency and then adding a feedback loop using an inverse dynamics controller to compensate for modeling errors. Theoretical analysis is presented to prove the convergence and stability of the proposed controller. Extensive experimental and comparison results also prove the superiority of the proposed controller over other algorithms.
Soft robotics, Modeling, Inverse dynamics
College of Engineering