A Recurrent Neural-Network-Based Real-Time Dynamic Model for Soft Continuum Manipulators

Abbas Tariverdi; Venkatasubramanian Kalpathy Venkiteswaran; Michiel Richter; Ole J Elle; Jim Tørresen; Kim Mathiassen; Sarthak Misra; Ørjan G Martinsen

doi:10.3389/frobt.2021.631303

A Recurrent Neural-Network-Based Real-Time Dynamic Model for Soft Continuum Manipulators

Front Robot AI. 2021 Mar 18:8:631303. doi: 10.3389/frobt.2021.631303. eCollection 2021.

Authors

Abbas Tariverdi¹, Venkatasubramanian Kalpathy Venkiteswaran², Michiel Richter², Ole J Elle^{3

4}, Jim Tørresen⁴, Kim Mathiassen⁵, Sarthak Misra^{2

6}, Ørjan G Martinsen^{1

7}

Affiliations

¹ Department of Physics, University of Oslo, Oslo, Norway.
² Department of Biomechanical Engineering, University of Twente, Enschede, Netherlands.
³ The Intervention Centre, Oslo University Hospital, Oslo, Norway.
⁴ Department of Informatics, University of Oslo, Oslo, Norway.
⁵ Department of Technology Systems, University of Oslo, Oslo, Norway.
⁶ Department of Biomedical Engineering, University of Groningen and University Medical Centre Groningen, Groningen, Netherlands.
⁷ Department of Clinical and Biomedical Engineering, Oslo University Hospital, Oslo, Norway.

Abstract

This paper introduces and validates a real-time dynamic predictive model based on a neural network approach for soft continuum manipulators. The presented model provides a real-time prediction framework using neural-network-based strategies and continuum mechanics principles. A time-space integration scheme is employed to discretize the continuous dynamics and decouple the dynamic equations for translation and rotation for each node of a soft continuum manipulator. Then the resulting architecture is used to develop distributed prediction algorithms using recurrent neural networks. The proposed RNN-based parallel predictive scheme does not rely on computationally intensive algorithms; therefore, it is useful in real-time applications. Furthermore, simulations are shown to illustrate the approach performance on soft continuum elastica, and the approach is also validated through an experiment on a magnetically-actuated soft continuum manipulator. The results demonstrate that the presented model can outperform classical modeling approaches such as the Cosserat rod model while also shows possibilities for being used in practice.

Keywords: Cosserat rod theory; Lie group variational integration; continuum manipulators; dynamic models; recurrent neural network; soft robotics.