Unknown System Dynamics Estimator for Motion Control of Nonlinear Robotic Systems

In this paper, we propose an alternative, simple, yet efficient estimation method to handle unknown dynamics and external disturbances for motion control of robotic systems. An unknown system dynamics estimator (USDE) is first proposed by introducing filter operations and simple algebraic calculatio...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on industrial electronics (1982) 2020-05, Vol.67 (5), p.3850-3859
Main Authors: Na, Jing, Jing, Baorui, Huang, Yingbo, Gao, Guanbin, Zhang, Chao
Format: Article
Language:English
Subjects:
Acceleration
Control stability
Control systems
Convergence
Coriolis force
Disturbance observers
Estimators
Mathematical analysis
Motion control
Motion stability
Nonlinear control
nonlinear disturbance observer (NDO)
Nonlinear systems
Robot control
robotic systems
Robotics
Service robots
System dynamics
Systems stability
Tracking
Tracking control
Tracking errors
Tuning
unknown system dynamics estimator (USDE)
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, we propose an alternative, simple, yet efficient estimation method to handle unknown dynamics and external disturbances for motion control of robotic systems. An unknown system dynamics estimator (USDE) is first proposed by introducing filter operations and simple algebraic calculations, where the external disturbances and unknown Coriolis/gravity dynamics can be estimated simultaneously. In the specific case where only the external disturbance is unknown, a further modified unknown disturbance estimator (MUDE) is introduced. These proposed USDE and MUDE can be easily implemented and their parameter tuning is straightforward compared with the nonlinear disturbance observer that requires calculation of the inverse of the inertia matrix. The acceleration signal of robotic joints is not used in the design of estimators. Moreover, we also show that the proposed estimators can be incorporated into the design of composite controllers to achieve satisfactory motion tracking response. The closed-loop control system stability and convergence of both the tracking error and estimation error are all guaranteed. Finally, the effectiveness of the two proposed methods is validated by using simulations and experiments based on a SCARA robot test-rig.
ISSN:0278-0046
1557-9948
DOI:10.1109/TIE.2019.2920604