Data-driven learning for robot control with unknown Jacobian
Unlike most control systems, kinematic uncertainty is present in robot control systems in addition to dynamic uncertainty. The use of different types of external sensors in various configurations also results in different sensory transformation or Jacobian matrices and thus leads to different kinema...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/152141 |
| _version_ | 1826112721177804800 |
|---|---|
| author | Lyu, Shangke Cheah, Chien Chern |
| author2 | School of Electrical and Electronic Engineering |
| author_facet | School of Electrical and Electronic Engineering Lyu, Shangke Cheah, Chien Chern |
| author_sort | Lyu, Shangke |
| collection | NTU |
| description | Unlike most control systems, kinematic uncertainty is present in robot control systems in addition to dynamic uncertainty. The use of different types of external sensors in various configurations also results in different sensory transformation or Jacobian matrices and thus leads to different kinematic models. Currently, there is no systematic theoretical framework in developing data-driven neural network (NN) learning and control methods for task-space tracking control of robots with unknown kinematics and dynamics. The existing NN controllers are limited to either dynamic control or kinematic control without considering the interaction between the inner control loop and the outer control loop. In this paper, a NN based data driven offline learning algorithm and an online learning controller are proposed, which are combined in a complementary way. The proposed task-space control algorithms can be implemented on robotic systems with closed control architecture by considering the interaction with the inner control loop. Theoretical analyses are presented to show the stability of the systems and experimental results are presented to illustrate the performance of the proposed learning algorithms. |
| first_indexed | 2024-10-01T03:11:45Z |
| format | Journal Article |
| id | ntu-10356/152141 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T03:11:45Z |
| publishDate | 2021 |
| record_format | dspace |
| spelling | ntu-10356/1521412021-09-13T07:26:50Z Data-driven learning for robot control with unknown Jacobian Lyu, Shangke Cheah, Chien Chern School of Electrical and Electronic Engineering Engineering::Electrical and electronic engineering Robot Control Neural Network Control Unlike most control systems, kinematic uncertainty is present in robot control systems in addition to dynamic uncertainty. The use of different types of external sensors in various configurations also results in different sensory transformation or Jacobian matrices and thus leads to different kinematic models. Currently, there is no systematic theoretical framework in developing data-driven neural network (NN) learning and control methods for task-space tracking control of robots with unknown kinematics and dynamics. The existing NN controllers are limited to either dynamic control or kinematic control without considering the interaction between the inner control loop and the outer control loop. In this paper, a NN based data driven offline learning algorithm and an online learning controller are proposed, which are combined in a complementary way. The proposed task-space control algorithms can be implemented on robotic systems with closed control architecture by considering the interaction with the inner control loop. Theoretical analyses are presented to show the stability of the systems and experimental results are presented to illustrate the performance of the proposed learning algorithms. Agency for Science, Technology and Research (A*STAR) This work was supported by the Agency For Science, Technology and Research of Singapore (A*STAR) , under the AME Individual Research Grants 2017 (Ref. A1883c0008). 2021-09-13T07:26:49Z 2021-09-13T07:26:49Z 2020 Journal Article Lyu, S. & Cheah, C. C. (2020). Data-driven learning for robot control with unknown Jacobian. Automatica, 120, 109120-. https://dx.doi.org/10.1016/j.automatica.2020.109120 0005-1098 https://hdl.handle.net/10356/152141 10.1016/j.automatica.2020.109120 2-s2.0-85087424196 120 109120 en A1883c008 Automatica © 2020 Elsevier Ltd. All rights reserved. |
| spellingShingle | Engineering::Electrical and electronic engineering Robot Control Neural Network Control Lyu, Shangke Cheah, Chien Chern Data-driven learning for robot control with unknown Jacobian |
| title | Data-driven learning for robot control with unknown Jacobian |
| title_full | Data-driven learning for robot control with unknown Jacobian |
| title_fullStr | Data-driven learning for robot control with unknown Jacobian |
| title_full_unstemmed | Data-driven learning for robot control with unknown Jacobian |
| title_short | Data-driven learning for robot control with unknown Jacobian |
| title_sort | data driven learning for robot control with unknown jacobian |
| topic | Engineering::Electrical and electronic engineering Robot Control Neural Network Control |
| url | https://hdl.handle.net/10356/152141 |
| work_keys_str_mv | AT lyushangke datadrivenlearningforrobotcontrolwithunknownjacobian AT cheahchienchern datadrivenlearningforrobotcontrolwithunknownjacobian |